Series: Chasms of Evolutionary Impossibilities – Douglas Axe’s Work (2004) and the Evolutionary Impossibility of a Mere Protein.

doi:10.1016/j.jmb.2004.06.058

8.3 “The Experiment Assumes Evolution in a Single Step”

When attacking a caricature — and ignoring what was actually tested

Objection

Some critics claim that Douglas Axe made a fundamental error in his 2004 study: he supposedly assumed that the evolution of complex systems happens in a single leap — as if a functional protein needed to emerge fully formed all at once. They argue that, in reality, evolution works through small, gradual steps, with intermediate forms that may have partial or different functions before reaching the final function.

🪜 For the lay reader: It is like saying we don't need to find a whole needle at once — we can find little pieces of it gradually. But Axe showed that even finding a single functional piece of the needle requires searching more haystacks than exist in the universe. And worse: these pieces only work if found together and perfectly fitted.

What Axe Actually Did

Axe did not ignore the idea of gradual evolution — he tested it directly. On page 1309 of the article published in the Journal of Molecular Biology, he explicitly discusses this possibility:

“Even if functional sequences were clustered in sequence space, the number of changes required to convert one cluster to another would present a formidable barrier.”

✅ Accessible methodological summary:

• Axe did not assume evolution occurs in a single leap
• He investigated whether gradual paths between functional sequences are truly accessible
• And discovered that even the shortest paths require too many coordinated changes to be feasible

🪜 Integrated analogy:

“It is like trying to cross a river by jumping from stone to stone. Axe asked: ‘Are these stones close enough for someone to jump?’ — and found they are not.”

Where is the Logical Error?

The criticism commits the straw man fallacy — attributing to Axe a position he never defended, only to refute it. Axe did not say evolution must happen in a single leap. He asked:

“Even if step by step, do we have enough steps? Enough time? Enough resources?”

🪜 Explanation for laypeople: It is like trying to assemble a 1,000-piece puzzle without knowing the final image, without knowing where the right pieces are, and without enough time to test all combinations. Even if you try to assemble it gradually, the chance of success remains absurdly small.

What the Data Show

Let's consider the most optimistic scenario possible for evolution:

• Global bacterial population: 10³⁰ cells
• Mutation rate: 10⁻⁹ per gene per generation
• Evolutionary time: 10¹² generations

Multiplying everything:

$$10^{30} \times 10^{-9} \times 10^{12} = 10^{33} \text{ possible attempts}$$

Now compare with the probability Axe found for the emergence of a functional protein:

$$P < 10^{-150}$$

🪜 Visual explanation: It is like having 10³³ lottery tickets — but needing 10¹⁵⁰ to have a real chance of winning. Even if you play every day, with all possible tickets, you don't even come close.

Model

Even if we needed only 5 coordinated mutations simultaneously — an extremely optimistic scenario — the probability would still be less than:

$$P \approx 10^{-150}$$

And even if we consider the entire universe as a supercomputer, with:

• Time since the Big Bang: 10¹⁷ seconds
• Atoms in the universe: 10⁸⁰
• Operations per second per atom: 10¹⁵

Total possible operations:

$$10^{17} \times 10^{80} \times 10^{15} = 10^{112}$$

Still, it doesn't even come close to the 10¹⁵⁰ needed.

🪜 For the lay reader: It is like trying to find a specific needle in a universe full of haystacks — and needing to find several pieces of the needle at the same time, which fit perfectly. And there aren't enough haystacks even in the entire universe.

What Does the Scientific Literature Say?

• Wolf-Ekkehard (2009): Shows that even two coordinated mutations require waiting time greater than the age of the Earth
• Lynch (2020): Admits that real populations are too small to accumulate multiple beneficial mutations
• Durrett & Schmidt (2008): Demonstrate mathematically that gradual evolution does not have enough time to work, even in fast-reproducing organisms

🪜 For the lay reader: These studies were not made by intelligent design advocates — they were made by evolutionists who recognize that the numbers don't add up.

Why This Criticism Fails

The criticism fails because it attacks a position Axe never defended — and ignores the calculations he actually made. Axe did not say evolution must happen in a single leap. He asked:

“Even if step by step, do we have enough steps? Enough time? Enough resources?”

And the answer, according to the data, is: no.

🪜 Final analogy:

“It is like trying to assemble a functional needle from pieces scattered across billions of haystacks — and needing all the right pieces to be in the same haystack, at the same time, and fit perfectly.”

Conclusion for the Lay Reader

Douglas Axe did not ignore gradual evolution — he tested it with real numbers, under generous conditions.

And discovered that even the shortest and most direct paths are beyond the resources of the universe.

The criticism makes a basic error: attacking a caricature of the study, not what was actually done.

Science is not made with assumptions — it is made with data. And the data show that functional evolution by random mutations, even gradual ones, is statistically impossible.

🪜 Visual summary:

“It's not that evolution needs to take a leap — it's that even small steps require stairs that don't exist.”

Therefore, this criticism does not invalidate the study.

It reinforces the strength of Axe's approach — and the depth of the problem faced by the traditional evolutionary model.

Priority Self-Refuting Sources (κ > 0.9)

• Wolf-Ekkehard (2009): Shows that waiting time for coordinated mutations exceeds the age of the Earth
• Lynch (2020): Admits that real populations lack sufficient scale to accumulate multiple mutations
• Durrett & Schmidt (2008): Demonstrate that even two coordinated evolutionary steps are statistically unviable