A Subway store manager wanted to invent a great new sub sandwich recipe – one that would make customers sing for joy and become regular patrons.
After considerable thought, he concocted an ingenious invention – a system for discovering a great recipe. Here’s how he did it.
He made a computer program to create random sub sandwich recipes, using random combinations of all the ingredients in the store. He offered customers to try the sandwiches for half price, as long as they were willing to score the sandwich afterwards.
After the customers rated the sandwiches, the computer program threw out the bad recipes and kept the good ones. It made small random changes to the ingredients in the “good” recipes, and created a new list of recipes. Then the new recipes were tried on customers.
This repeated over several weeks. Gradually, the recipes got better, and customers scored them higher and higher. Eventually, the store manager was very happy to pick a winning recipe. The new sub sandwich went on sale, marketed as the “Subway Evolution”.
So, was this evolution? Well, this is a fast-food example of “genetic programming” that has been used in engineering fields, such as for designing a good digital filter. It contains several aspects of evolution. It has “genetic variation” of a sort. It has natural selection and survival of the fittest. Hooray – it’s evolution!
But it has limitations. For example, would the manager’s system ever produce a cake recipe? No, because the potential ingredients were limited to the store’s sub sandwich ingredients list. Would it ever invent a delicious new sauce? Not possible.
This system is a practical allegory of evolution with hard boundaries – a system that can produce the “fittest” sub sandwich but that’s all.
The question is, does the natural world work like this, or can evolution continue without bounds? Darwin observed variation in finch populations as conditions changed. Evolutionists say evolution is that process just carried on over longer periods of time. But creationists say the variations he saw were within the normal scope of genetic variation in the population. The finches always had a range of beak sizes in their genetic scope, and the varying weather conditions “selected” some beak sizes as being fitter for given conditions. Furthermore, were genetic mutations needed to vary the average beak sizes? It’s more likely that only the normal genetics of sexual reproduction were involved.
Evolutionists will argue that macroevolution is certainly possible and observed, by defining macroevolution as anything that produces “speciation” – the appearance of a new species. But that is just dodging the heart of the issue on a technicality. The definition and cause of “speciation” is complicated. The Bible talks about animals being made “after their kind” (Genesis 1:24). A “kind” could be a broader category than “species”, perhaps representing the complete genetic scope for variation through breeding in a population of animals. For example, tigers and lions are distinct species, but have demonstrated the ability to interbreed. Perhaps many species of big cat fit into one Biblical “kind”. So while a population of animals could achieve what evolutionists call “speciation”, the extent of variation could still be limited by what the Bible calls “kind”.
So as a creationist, I say this: God created animals with scope for genetic variation. But that variation has boundaries, and animals can’t go beyond those boundaries. No matter how much time you give it, the store manager’s system will never give you a chocolate cake recipe. No matter how much time you give it, Darwin’s finches will never turn into lizards.
But what about mutations? Don’t they provide genuine “new genetic material”, allowing a population to move beyond the bounds of what it can achieve through sexual reproduction? In the example of the Subway store, perhaps the computer could have a “glitch” thanks to a cosmic ray. Then instead of listing “mustard” it might ask for “mustare” or even better, “bustard”. As an embedded software engineer, my experience is that “glitches” are disastrous. At best, the software recovers gracefully. More likely it crashes so you have to cycle the power to get things going again. So in the world of genetics, are mutations any better, or do they just “mess things up”? As far as I can tell, genetic mutations are a disaster too. In many cases, the damage is “repaired” by the cell machinery. In some cases, the damage causes genetic diseases. But in no cases does a mutation create a genuine “new feature”.
In summary, sub sandwiches will always be sandwiches. Finches will always be finches.