« How To Fail With External DSLs | Main | Canalizing Design »

April 10, 2009


Marcel Popescu

Ignoring the evolution part for now - do you know WHY you believe that the Earth revolves around the Sun, and not the other way around? I find that an amazingly educational question when dealing with evolutionists.

Jim Sweeney

Observations of independent organisms working together as one doesn't begin to explain the complexities of life. You're not looking at inorganic material that somehow independently begins to follow rules. The parts themselves are alive and inexplicably complex and the simulated computer cells need to be infused with instructions from their...er, creator.

Michael Feathers

Jim: Inorganic matter follows rules all the time (Newton's laws of motion, etc). It doesn't need any special intelligence to do so. In Conway's Life, the "creator" just creates the ground state. The state of the world and the rules determine what happens next. If you feed enough energy, time, and chaos into a local system, it starts to organize.

Dave Hoover

I agree with Jim Sweeney on the creator role in Conway's Life. Aren't I, as the simulator runner, creating something which then takes on a life of its own? It seems to me that Conway's Law depends on the creation of something by an intelligent being and then simulates the evolution of that creation.

Michael Feathers

Dave: Fair enough, but I don't think that implies anything in particular about the universe. One the one hand, we can have an infinite regress of creators or we can consider that the notion of creation itself is a cognitive bias: we identify agents in the world and phenomena that we consider distinct enough from our scale and point of view to perceive as beginnings. It's not clear to me, though, that the concepts of "beginning" and "creation" have meaning independent of our perception.

Patrick Wilson-Welsh

Michael: insightful, thoughtful post as usual. Thanks. All: crazy suggestion: faith in spirit (insert your God-synonym here) is an internal, largely somatic experience, and cannot be independently corroborated. Therefore, we cannot usefully debate it (for more on this, see Ken Wilbur's writings). So, I'll repurpose Laurie Anderson's saying about music here: "Talking about music [faith] is like dancing about architecture."

The best we can hope for is to honor each other's perspective's and internal experience of reality, and learn from them. Spirit is the opposite of empirical science.


life coming from non-life (biogenesis) isn't part of evolutionary theory, as far as I know

Michael Feathers

mgroves: I always just assume abiogenesis, I guess. The way I see it, nature doesn't do crisp distinctions well. Is a virus alive or not? It really depends on the definitions we choose to use and I suspect that as we learn more the life/non-life distinction will look as arbitrary as the definition of a planet. To me, all of it, life and non-life is a vast pattern of matter and energy. In everyday speaking we use the words life and non-life, but they are labels: the map not the territory. I think I've internalized that enough to just assume that abiogenesis is trivially true because we set the boundaries.

I agree that evolution and abiogenesis are distinct concepts, but I do think that man-in-the-street objections to them do coincide.


Speaking of Earth revolving around the Sun, you *do* know that it's not true, right? I mean, the relativity principle clearly says that there's no absolute point of view. Yet, we all believe in the Sun being "still" and the Earth moving around.

Why? I think it's because relativity is hard to grasp. We believe in absolute things. Absolute "exists" or "doesn't exist". And this is where all holy wars begin (note that an irritation is a minor internal holy war).

What I propose is a quantum view (and I'm not the first in that). Whether a god or several gods exist or not is a question similar to whether an electron is a particle or a wave. Let's say it depends on the observer, and the holy war ends.

Note that this doesn't contradict any belief: I can still believe in whatever my guts (or soul) tells me. As a matter of fact, I do believe in gods *and* Darvin's theory, at the same time realizing that it's just my personal belief and not the absolute truth.

Jim Sweeney

I didn't say that inanimate matter does not obey rules, but rather that it does not independently begin to follow rules based on the random generation of new information. The addition of organizing information to a system requires an external intelligence.

Michael Feathers

jim: I'm having trouble parsing that. I don't think there's any such thing as inanimate matter. We never get all the way to absolute zero. Or, are you drawing another distinction? Do you see a mouse made of metal and semiconductors as less animate than one made of amino acids? If so, why?

Also, what is "organizing information"? Do you see it as different than other information? What makes it special?

Keith Braithwaite

If you liked this, you might also like "Autopoiesis and Cognition" by Maturana and Varela.

Jim Sweeney

We can quibble about semantics, but I think we all have pretty good intuition about the qualities of a living organism vs. non-living matter. Let's not pretend there is any genuine confusion on that matter. We should also understand that the simplest most rudimentary life forms contain a tremendous amount of encoded information that allow biological processes to continue and be passed on and that the sheer quantity and organization of this information makes it mathematically impossible for this data to have come together by chance even given the most far reaching estimates for the age of the universe. When we add the lack of any observable inorganic process adding such information to organic systems, it becomes an open and shut case debunking macroevolution as an explanation for life and the variety of life that we see today.

Michael Feathers

jim: With respect, "semantics" is all there is. We have to agree on meanings to communicate. I agree that a mouse and a mechanical mouse are different in many ways, but the question is whether they are different in essential ways. Is, for instance, a virus alive? As in most things, we learn by looking at the interesting cases rather than relying on gut feel.

BTW, this isn't true: "it is mathematically impossible for this data to have come together by chance even given the most far reaching estimates for the age of the universe." First off, statistical phenomena are not impossible, they are just more or less likely. Secondly, if you're going to say something like that, it would be good to justify it.

Jim Sweeney

In this context, requiring that we agree on a precise definition of life just clouds the issue. The point is that those things that reasonable people might agree are alive (including viruses) are too complex to have arisen by chance and cannot be observed to evolve (become more complex) by the random generation and infusion of new information

I'll withdraw my phrase "mathematically impossible" and replace it with "so statistically improbable that no honest mathematician would give the premise any serious consideration". I understand that those whose religious convictions do not allow them to consider simpler solutions must persist in their "faith", but hopefully they will not blind themselves to the vast disparity between man-made computer models and biological systems observed in nature.

Now, having annoyed you, I hope you will allow me to compliment you on your terrific work on Working Effectively with Legacy Code. It has been a tremendous help in understanding how bad code can be (dare I say it?) evolved into a testable, well-organized system. Maybe a better application of Conway's Life would be to consider how it suggests that software building processes might be improved and that systems allowed to evolve naturally can do so in an orderly way simply by starting with a better set of principles.

Michael Feathers

jim: I think you have to justify this. I don't see any evidence that it is true: " The point is that those things that reasonable people might agree are alive (including viruses) are too complex to have arisen by chance and cannot be observed to evolve (become more complex) by the random generation and infusion of new information."

Thanks for the WELC compliment. I'm not annoyed. I look at emergent design as being loosely analogous to what you're speaking of. If we follow Beck's rules of simple design, we can end up with very good software but it is not fully deterministic. The better outcomes need guidance, much like if I were designing humans I'd think twice before passing the birth canal through the pelvis, female reproductive structures which lead to ectopic pregnancies, or having the testes of males develop in the abdomen leading to hernias etc.

Seriously, though, let me throw something past you. Suppose there was a designer. Would you be okay with the idea that that designer had a designer too?

Jim Sweeney

Touching on your points in reverse order: There is no need for an infinite regression of creators if the creator is himself infinite and without beginning. Mankind has a fundamental understanding of this as we observe that the ancient traditions of most people groups include the notion of an "uncreated creator". Read "Eternity in Their Hearts" by Don Richardson.

I haven't read Bain's Emergent Design yet, but it's on my Amazon wish list. I love the notion of a software system growing "organically" but yes, even the most sound design principles instilled in the beginning do not remove the need for the constant inspection and intelligent guidance that an Agile process can afford.

I wasn't attempting to make this into a theological discussion, but I can only address some of the points you raise such as apparent mistakes in the human design, by using the Bible. I'll understand if you don't want to go there.

Finally, you say you don't see any evidence in support of my negative statements about complex biological systems arising by chance. This is essentially asking me to prove a negative. One can't provide universal evidence that an event has NOT happened or does not happen. It is rather the responsibility of the evolutionist to demonstrate that natural random processes CAN be observed to add information to a system, increasing its functional complexity. Anywhere. Ever.

Michael Feathers

jim: there are religious traditions which see the universe as infinite in both temporal directions also - a never ending cycle. I asked the question about a designed designer because I suspect that it doesn't sit well with many religious people. A designed designer is somehow less special than one which stands outside of time. It's worth asking why, though. And, why as well that the stuff of the universe can't sit outside of time in the same way that a designer can.

Re proving a negative, all I'm saying is that you postulated a negative. As the person who did, it would be great if you back it up.

Re adding information.. I think that, again, it is going to have to come back to definitions. I know from your comments that you program, so how about this explanation: imagine a grid in Conway's Life where you randomly turn cells on and off. At some point in time, randomly, a 5-cell configuration of cells known as a "glider" appears (google it). Gliders are more structured than most random configurations. They move, and they have discernable qualities, all of which count as information by my definition. There's an field of mathematics called 'random graphs' which is entirely about showing that when organized forms are inevitable given a certain number of elements.

Are we talking about the same kind of information?

Stan Kurdziel

One of the things I have pondered is that assuming single cellular life exists, why would independent single cellular organisms organize into cohesive multi-cellular units?
I like your comparison to a flock of birds and Conway's game of life. Perhaps for certain configurations, the symbiosis between the single cells becomes so strong that they are no longer independent. Doesn't seem sound so far fetched when phrased that way.

Simple rules leading to complex behavior also reminds me of some of A New Kind of Science by Stephen Wolfram: http://www.wolframscience.com/
Of course, that book is too big to read the whole thing! =P

Stan Kurdziel

Also, check out this video by Craig Venter on creating synthetic life: http://www.ted.com/index.php/talks/craig_venter_is_on_the_verge_of_creating_synthetic_life.html
Pretty fascinating.

Jim Sweeney

I would expect most religious traditions to intuit that time extends indefinitely in both directions, but it need not be cyclical. The idea of a created creator is not unheard of among religious people. Mormons believe that Elohim had a creator. The notion doesn't sit well with me because this is not how the Bible describes God. There is also the argument that a causal chain cannot be of infinite length.

Yes, I think there is some gap between our uses of the term "information". I'm talking about the kind of information that needs to be added to a fish's DNA in order for its descendants to "progress" toward growing legs. What you are describing I would say is the illusion of that kind of information in the same way that a.i. programs that talk to you are displaying the illusion of intelligence. Both are products of genuine intelligence providing the information and the environment necessary to support the illusion.

In scientific terms, "postulating a negative" is something of an oxymoron. Let's say I'm making a claim that can't be proved, but on the topic of finding evidence that natural processes ever add information to say, an organism's genetic code, the burden ought to be on the person claiming that this happens. If I'm holding a gold nugget and I tell you that I created it from lead a moment ago, you will naturally doubt my word and ask me for evidence. If I can't produce any, you will be entirely justified in stating a negative, claiming "there is no evidence that lead turns to gold" and you ought to feel no burden to support this claim. Atheists make negative statements about the existence of God without backing them up all the time. From their perspective, the burden of proof is on believers and this is quite a reasonable position. You can't prove a negative. Unfortunately, from the believer's perspective, they seem to be ignoring quite a lot of evidence.

Luca Minudel

there are connections between complexity science, social application on complexity science, agile team dynamics and Conway's Life agent based simulation
the same apply to the heuristic approach on agile sw development

the post "Master Craftsman Teams" of Uncle Bob seems to me out of sync with this - or I just misunderstood that post completely


Interesting post. I'm of the same camp as Jim Sweeney. The one thing that all this boils down to is either there was an eternal Being that created all things, or eternal matter and energy. The question where did either come from, can't be answered, so a designer of a designer just postpones the question of what to put faith in. The one thing in your opening post, I would say isn't true of all Creationists or IDers, is that the universse is centered around man. As a Christian, I believe everything is created for God's glory and man plays a central role in that creation, but he isn't the focus.

Dr. Bubba

"Evolution, on the other hand, is a tough one. The time scales are huge..." - MF

That could be because evolution can be considered part of the forensic sciences where we have to extrapolate more than interpolate. Extrapolate is a harder problem with less certainty in many cases.


I find it interesting that undirected evolution of complex lifeforms is supposed to have some similarities to Conway's game of Life. Already mentioned, Conway's Life requires an intelligent agent to set up the scenario it will run. What hasn't yet been mentioned is that the universe that runs these rules against the input data was created and designed by an intelligent agent, a programmer. It doesn't seem too much of a stretch to see that the intelligently created rules of Life, simple as they are, can create interesting and complex results, and perhaps equally so, there are simple, yet created, rules that governs events in nature. Yet it seems incredibly unintuitive to suppose that something even as simple as the software to run the algorithm Life is based on could somehow form in an undirected, unguided manner.

As it turns out, a sizable portion of intelligent design supporters come from software and engineering disciplines because they recognize how difficult it is to intelligently create complex systems, and none of those fields produce anything remotely as complex as even the simplest cell.

The comments to this entry are closed.