Can Evolution and Intelligent Design Both Be Tested Scientifically?8 min read

Critics of intelligent design claim it is merely creationism in disguise and inherently untestable, but both assertions are incorrect.

Creationism begins with the authority of the Bible, whereas intelligent design starts with observations in nature, remaining open to the possibility that complex functional information could originate from an intelligent source rather than being restricted to natural processes alone.

To delve deeper, while the scientific method employs methodological naturalism—seeking natural explanations for phenomena—our conclusions need not be confined to philosophical naturalism, which assumes only natural processes can account for all phenomena. This distinction sets the stage for evaluating both evolution and intelligent design as frameworks that employ scientific reasoning, including empirical and inferential approaches, to explain the origin and development of life. This article argues that both make testable, falsifiable predictions and should be evaluated by the same methodological standards.

1. Empirical and Inferential Science

Empirical science involves repeatable experiments and direct observation. Inferential science, by contrast, involves reconstructing the past from physical evidence when direct observation is not possible. For more see The Three Types of Science: Experimental, Inferential, Fantasy (wholereason.com) ¹

1.1 Inferential Forensics vs. Empirical – An Example

A classic example of inferential science is a criminal investigation. We cannot observe the murder as it happened, but we infer what occurred from fingerprints, DNA, blood patterns, and the state of the body.

Sometimes, empirical methods can be applied—for example, by testing rates of decay to estimate time of death. Similarly, intelligent design and evolution use inferential reasoning to interpret historical evidence, supplemented by empirical tests where possible.

“Scientific investigation often involves using present processes and data to infer past causes.” — Stephen C. Meyer, Signature in the Cell (HarperOne, 2009)

However, even though we are investigating a non-repeatable historical event which cannot be directly observed, if there are ongoing physical or biological processes still occurring since the incident, we can use those to run empirical tests, such as measuring how much a body has decayed. These tests provide “testable” data, but are still historical evidence, and do not allow direct observation of the event itself.

2. How Evolution Is Studied

2.1 Ongoing Evolution: Empirical Studies

Because evolution is assumed to be an ongoing process, it can be subjected to empirical tests. One of the most well-known experiments in evolutionary biology is the Long-Term Evolution Experiment (LTEE) conducted by Richard Lenski.

The LTEE showed adaptations in E. coli over thousands of generations. However, these adaptations universally involved loss-of-function mutations, not the emergence of novel genes or structures. For a detailed critique of that landmark experiment, see A Critical Examination of the Lenski Long-Term Evolution Experiment: Loss of Function, Not Novelty (wholereason.com) ²

2.2 Historical Evolution: Inference from Fossils

Fossil evidence does not allow us to directly observe one species turning into another. We infer relationships from the fossil record, anatomical similarity, and dating methods, making this an inferential process.

“No one has ever observed the transformations required by macroevolution. They are inferred from patterns.” — Henry Gee, In Search of Deep Time (Free Press, 1999)

3. How Intelligent Design Is Studied

3.1 Empirical Tests for Design Predictions

Intelligent design proposes predictions that can be empirically tested. For example, it predicts that genetic information will not arise by purely unguided processes and that biological systems will exhibit irreducible complexity. Additionally, it expects genomic function to be widespread rather than accidental or “junk.”

“If an intelligent cause designed life, we would expect most DNA to be functional. The prediction that ‘junk DNA’ is mostly functional has been confirmed by ENCODE.” — Jonathan Wells, The Myth of Junk DNA (Discovery Institute Press, 2011)

3.2 Inferential Investigation of Design

Like forensicadj. employing scientific analysis of physical evidence (as from a crime scene) More science or cosmology, intelligent design relies on inferential reasoning, looking for signs of specified complexity, purposeful arrangement, and functional coherence—features typically associated with intelligence.

“We have repeated experience of rational agents producing specified complexity.” — William Dembski, The Design Inference (Cambridge University Press, 1998)

Dembski formalizes this by proposing a “universal probability bound.” If the probability of a specified event occurring by chance falls below this bound, it is rational to infer design.

“When an event is specified and its probability under the chance hypothesis is sufficiently small, then we can eliminate chance and infer design.” — William Dembski, No Free Lunch (Rowman & Littlefield, 2002)

4. Contrasting Predictions of Evolution and Design

4.1 DNA Functionality

Evolutionary theory has historically predicted that, due to mutational load, a significant portion of the genome should be non-functional—so-called “junk DNA.” This stems from the idea that random mutations accumulate over time, and natural selection cannot efficiently eliminate non-functional sequences. Some evolutionary biologists estimate that no more than 10-20% of the human genome is functional due to constraints imposed by mutational load.

“Given the rate of deleterious mutations, it is unlikely that more than 10-20% of the genome can be functional, as the mutational load would otherwise be unsustainable.” — Dan Graur et al., “On the immortality of television sets: ‘function’ in the human genome according to the evolution-free gospel of ENCODE,” Genome Biology and Evolution (2013)

By contrast, intelligent design predicts that most, if not all, of the genome will have function, as a purposefully designed system would minimize non-functional components. Findings from the ENCODE project, which suggest over 80% of the human genome exhibits biochemical function, support this prediction and challenge the notion of widespread “junk DNA.”

“The ENCODE project has revealed that at least 80% of the human genome is associated with some form of biochemical activity, suggesting pervasive functionality.” — The ENCODE Project Consortium, “An integrated encyclopedia of DNA elements in the human genome,” Nature (2012)

4.2 Genetic Degradation vs. Novel Gain

Evolution requires novel genetic features to emerge through unguided processes. However, empirical observations, such as those in Richard Lenski’s Long-Term Evolution Experiment (LTEE), often show genetic reduction or loss-of-function changes that improve fitness by simplifying existing systems. The LTEE, initiated in 1988 with 12 populations of E. coli and running for over 60,000 generations, has primarily demonstrated adaptations through loss-of-function mutations. For example, the emergence of aerobic citrate metabolism in one population was not due to a novel ability but the loss of regulatory mechanisms that suppress citrate metabolism in aerobic conditions. (A Critical Examination of the Lenski Long-Term Evolution Experiment: Loss of Function, Not Novelty (wholereason.com) ³

“The citrate-using phenotype did not involve the evolution of a new function but rather the loss of a regulatory mechanism that represses citrate transport under aerobic conditions.” — Michael Behe, Darwin Devolves (HarperOne, 2019)

Intelligent design, while not denying adaptation, predicts that unguided mutations will not generate fundamentally new genetic information. The LTEE’s findings align with this, as no novel genes or complex systems emerged over 60,000 generations, only modifications or losses of pre-existing functions.

4.3 Falsifiability Through Experiment

Both theories offer falsifiable predictions. For instance, if Lenski’s E. coli had shown a novel gene or system arising that was clearly not the product of broken or modified pre-existing genes, it would challenge the intelligent design framework.

“A single example of a complex system arising by unguided means would falsify ID.” — Michael Behe, Darwin’s Black Box (Free Press, 1996)

5. Conclusion

Far from being unscientific, both evolution and intelligent design employ empirical testing and inferential reasoning. Both make testable, falsifiable predictions and can be evaluated based on evidence. Science is not about what you can observe today but what you can explain using reliable methods of reasoning and experimentation. By distinguishing methodological naturalism from philosophical naturalism, we see that intelligent design and evolution belong to the same scientific conversation.

“Science is not defined by its conclusions, but by its method of inquiry.” — Thomas Kuhn, The Structure of Scientific Revolutions (University of Chicago Press, 1962)