Argument

The Living Source Argument

Intro

Life began. Something caused it. The question is whether that something was an undirected chemical process or an agent.

Every pathway from raw chemistry to the first cell faces the same stack of problems: amino acids have to be left-handed, they have to be concentrated rather than diluted, the sequences have to carry functional information, and you need both the storage system and the reading machinery at the same time. Seventy years of origin-of-life chemistry have not solved any of these jointly. The best explanation for coded, functional information at biological scale is an intelligent agent.

But that raises a second question. If the agent that made the first life was itself embodied, made of coded functional information in a physical system, then the same argument applies to it. Where did its information come from? The universe's past is finite, so the chain of embodied agents cannot run forever. It ends in an agent that is not embodied: personal, immaterial, living, and powerful. That is a central part of what is meant by God.

◎ See this page's graph and discover linked topics →

In full

The Living Source Argument is a three-stage regress from the origin of biological life to an unembodied intelligent source. Stage 1 establishes that life began and that an undirected chemical process is not the best explanation for its cause, leaving an intelligent agent. Stage 2 applies the same information-needs-an-agent principle to any embodied agent (aliens, embodied simulators, directed panspermia sources), generating a regress that the universe's finite past terminates. Stage 3 identifies the terminus: an agent that is personal (agency is personal), immaterial (not built from information-bearing matter), living (the source of life), and powerful (the cause of the first life and of every embodied agent in the chain). The logic is deductively valid. The strength of the conclusion depends on Premise 5 (the information-to-agent inference), which is an inference to the best explanation, not a demonstration. The argument does not by itself show that God is one, necessary, or eternal; the Contingency Argument and a simplicity argument supply those in a cumulative case. The argument answers the standard escape routes: aliens and embodied simulators fall into the Stage 2 regress, natural panspermia still faces the information problem, and the "who made God?" objection fails because the regress premises apply only to embodied systems built from information.

Format note

A braided case runs four kinds of reasoning on one argument instead of choosing one. A deductive spine of numbered premises carries the conclusion, so the argument is valid and every disagreement is forced onto a named premise. Each premise is then held up by an inductive strand (the data that make it probable), an abductive strand (rival explanations of that premise's data, compared on explanatory power, scope, simplicity and how ad hoc they are) and a convergence note (independent disciplines that reach the same premise). A knot checks that the strands are genuinely independent, so no datum is counted twice. A likelihood ledger states, datum by datum, how expected the evidence is on theism compared to its rival, as honest qualitative ranges. A defeater gate answers the single strongest objection to the whole argument before the cheatsheet.

The spine

Validity. Stage 1 (P1-P6) is a disjunctive syllogism: life began, it was caused, the cause was either undirected process or agent, the undirected process is ruled out, so it was an agent. Stage 2 (P7-P9) applies P5's principle to any embodied agent, generating a regress. Stage 3 (P10-P11) closes the regress by the finitude of the past. C draws out the properties of the terminus.

Strength of the conclusion. The deductive structure is valid. The empirical weight falls on P4 and P5, which rest on inference to the best explanation. Call the conclusion "best explanation," not "proof," in debate.

P1, Life began to exist

Inductive strand

1. The fossil record. The earliest plausible microfossils date to roughly 3.5 billion years ago. Before that, Earth's surface was molten. Life does not appear in the geological record before a certain point.
2. The age of the Earth. Radiometric dating converges on approximately 4.5 billion years. The universe itself is approximately 13.8 billion years old. Neither is eternal.
3. Thermodynamic arrow. The second law of thermodynamics implies that the universe began in a low-entropy state and is running down. An eternal past with biological life is thermodynamically untenable.

Abductive strand

Convergence

Geology (fossil record), physics (radiometric dating, thermodynamics), and biology (absence of life before ~3.8 Ga) all converge on P1.

Objections and rebuttals

1. "Life could have existed in forms that leave no fossil trace." Rebuttal. Even granting this, the finite age of the Earth and the universe means life began at some point. The claim is about temporal beginning, not about earliest detectable record.

P2-P3, The cause was either undirected process or agent

Inductive strand

1. The principle of sufficient reason. Events that begin to exist have causes; this is foundational to every science.
2. The disjunction is exhaustive. The cause of a new phenomenon is either directed (by an agent with intentions) or undirected (by natural law and chance operating without intention). There is no third option.

Objections and rebuttals

1. "Quantum events are uncaused." Rebuttal. Quantum indeterminacy means unpredictable, not uncaused. The quantum vacuum is a structured state governed by laws, not nothing. And even granting genuinely uncaused quantum events, the origin of life involves macroscopic chemistry, not single quantum transitions.

P4-P5, Undirected chemistry fails; agency is the best explanation

Inductive strand

1. Chirality. Biological amino acids are exclusively left-handed. Undirected chemistry produces a racemic (50/50) mixture. No known natural mechanism selects for homochirality at the required purity.
2. The dilution problem. Prebiotic chemistry in an ocean dilutes reactants below useful concentration. Mineral surfaces and tidal pools help locally but introduce new problems (surface bonding competes with polymerisation).
3. The sequence problem. A functional protein requires a specific sequence of amino acids. Douglas Axe estimated that roughly 1 in 10^77 random 150-residue sequences folds into a functional protein (Journal of Molecular Biology, 2004). The search space is astronomically larger than what chance can explore.
4. The chicken-and-egg problem. DNA stores information but needs protein machinery to read it. Protein machinery is built from information stored in DNA. Neither works without the other. The RNA-world hypothesis attempts to solve this with a single molecule that stores and catalyses, but ribozymes are fragile, short-lived, and have not demonstrated self-replication at biological fidelity.
5. Pasteur's principle. In 160 years of microbiology, life has been observed coming only from pre-existing life. See Biogenesis Argument.
6. The uniform experience with information. Computer code, written language, engineering blueprints, and archaeological artifacts all trace to minds. No undirected natural process has been observed producing coded, functional information at biological scale. See Argument from Origin of Life.

Abductive strand

Convergence

Synthetic chemistry (Tour: "we have no idea how to make a cell"), information theory (Yockey), probability (Hoyle, Koonin), and the historical uniformity of biogenesis (Pasteur) converge on the inadequacy of undirected chemistry.

Objections and rebuttals

1. "Give science more time." Rebuttal. The argument is not "science cannot explain it, therefore God." It is that the kind of thing required (coded functional information) is the kind of thing intelligent agents produce and is not the kind of thing undirected chemistry has ever been observed producing. The inference is to the best current explanation, which is how historical sciences work. A promissory note is not an explanation.
2. "Lab progress shows it can be done." Rebuttal. Miller-Urey, Sutherland's ribonucleotides, Szostak's protocells, and self-replicating ribozymes show what skilled chemists can do with purified reagents and staged steps. They show what intelligence produces, not what unguided chemistry produces. See Lab-vs-Wild Defeater for Origin-of-Life Demonstrations.
3. "We just need a simpler start." Rebuttal. The minimal cell experiments (Venter's JCVI-syn3.0, 473 genes) show that the floor for a self-sustaining cell is high. Below that floor, you do not have something simpler that also works; you have something that does not work. See Minimal Cell.

P7-P9, Embodied agents need prior agents (the regress)

Inductive strand

1. Aliens and extraterrestrial designers. If the first life on Earth was seeded by an extraterrestrial civilisation, those aliens are themselves embodied organisms built from coded, functional information. The same argument applies to them. Where did their information come from?
2. Embodied simulators. If we live in a simulation run by embodied beings, those beings are themselves composed of information-bearing matter. The regress applies.
3. Natural panspermia. If life arrived on a meteorite from elsewhere, the chemistry is the same everywhere. Chirality, dilution, sequence specificity, and the chicken-and-egg problem are not solved by moving the location. And the universe's past is finite.

Abductive strand

Objections and rebuttals

1. "Maybe aliens seeded life here." Rebuttal. Granted, but that pushes the question back one step. Where did the aliens' coded information come from? If they are embodied, the regress applies to them. The argument does not deny that intermediaries exist; it denies that embodied intermediaries can be the ultimate explanation.
2. "Maybe we live in a simulation." Rebuttal. If the simulator is embodied, the regress applies. If the simulator is not embodied, you have arrived at the conclusion.

P10-P11, The finite past closes the regress

Inductive strand

1. The Borde-Guth-Vilenkin theorem (2003). Any universe that has, on average, been expanding throughout its history must have a past boundary. This applies even to inflationary and multiverse scenarios. (Cited from memory; verify before live deployment.)
2. The second law of thermodynamics. A universe with a finite supply of usable energy that has been running down cannot have been running forever. An actually infinite past would have reached heat death already.
3. The impossibility of traversing an actual infinite. An infinite series of past embodied-agent-causes, each depending on the one before it, would require that an actually infinite number of events had already been completed to reach the present. This is the Hilbert's Hotel family of arguments. See Kalam Cosmological Argument.

Objections and rebuttals

1. "The Borde-Guth-Vilenkin theorem has exceptions." Rebuttal. The theorem applies to any spacetime with average expansion greater than zero. The exceptions (contracting universes, precisely static models) are not empirically supported. The second law provides an independent line to the same conclusion.
2. "Who made God?" Rebuttal. The regress premises (P7-P8) apply only to embodied systems built from coded, functional information. P2 applies only to things that begin to exist. God is not embodied and did not begin, so neither premise reaches Him. The question "who made God?" is like asking "who married the bachelor?" It misapplies a category.

The knot

The strands are independent. The chemistry data behind P4-P5 (chirality, dilution, sequence odds, chicken-and-egg) come from laboratory science and probability theory. The regress in P7-P9 is a logical move applying the same principle consistently. The finitude of the past in P10 comes from cosmology and thermodynamics. No datum is counted twice: the chemistry establishes that an agent is needed, the regress establishes that an embodied agent is not enough, and the cosmology closes the chain.

The ledger

Prior-agnostic: the ledger does not assume theism is likely or unlikely in advance. It asks only how expected each datum is under each hypothesis.

Net. The rows jointly make the evidence significantly more expected on theism than on naturalism. The strongest single row is the coded-information datum. The argument does not claim to prove God with mathematical certainty; it claims the evidence is far better explained by an intelligent, unembodied source than by undirected chemistry and an infinite chain.

The defeater gate

The strongest single objection: "Minds without bodies are impossible. There is no evidence of disembodied intelligence."

Steel-manned: every mind we have ever observed is embodied in a brain. Positing a disembodied mind is an extraordinary claim requiring extraordinary evidence. The concept may not even be coherent.

Reply. The objector now faces a dilemma.

1. If minds require bodies, and bodies require coded functional information, and coded functional information requires a mind (P5 and P8), then the first mind had no body, which contradicts the premise. The regress forces a disembodied starting point whether or not the objector likes it.
2. If the objector drops P5 or P8 to avoid the conclusion, he must explain where coded functional information comes from without an intelligent source. But dropping P8 also undercuts the design inference that SETI, archaeology, and forensics rely on. The principle "coded information implies a mind" is not an apologetic invention; it is the foundation of information science.
3. The claim "minds without bodies are impossible" is an empirical generalisation from a limited sample (known organisms). It cannot be elevated to a metaphysical necessity without assuming materialism, which is the very question at issue.

The defeater does not prove disembodied minds exist. It shows that the objector's own premises, applied consistently, require one.

Concessions

• Grant: P5 is an inference to the best explanation, not a logical demonstration. Now collect: inference to the best explanation is how all historical sciences work, including the naturalist's own account of evolution, so the method is not special pleading.
• Grant: the argument does not show that God is one, necessary, or eternal. Now collect: those properties are supplied by the Contingency Argument and a simplicity argument. This argument establishes that the source of life is personal, immaterial, living, and powerful, which is a substantial conclusion on its own.
• Grant: future science could discover a naturalistic pathway to life. Now collect: the argument runs on current evidence and current best explanation, which is how historical reasoning always works. A promissory note that a future discovery will change the picture is not a present defeater.

Cheatsheet

The 30-second version (read aloud).

Life began. Something caused it. Seventy years of chemistry have not found a way for unguided processes to build a cell, because a cell runs on coded information, and the only source of coded information anyone has ever found is a mind. So the first life points to an intelligent agent. But if that agent had a body, it was built from coded information too, and you need another agent behind it. The universe's past is finite, so the chain ends in an agent that does not have a body: personal, immaterial, living, powerful. Call that what you want. I call it God.

Fast facts:

1. The simplest free-living cell needs roughly 470 genes, hundreds of proteins, and a code to read them. No undirected chemistry has produced anything close.
2. Douglas Axe estimated that roughly 1 in 10^77 random 150-residue sequences folds into a functional protein.
3. Aliens and simulators do not escape the argument; if they are embodied, the regress applies to them.
4. The Borde-Guth-Vilenkin theorem (2003) shows that any universe with average expansion greater than zero has a past boundary.

The closing line.

Life is not an accident. It is a message. And a message always has an author.

Cross-examination

Round 1, the information question

• Q: Has any undirected chemical process ever been observed producing coded, functional information at biological scale?
• If they say: "Not yet, but give it time." Then: A promissory note is not an explanation. We go with the best current explanation, and the best current explanation is agency.
• If they say: "RNA can self-replicate." Then: Show me a ribozyme that copies itself at biological fidelity without a researcher designing the experiment. Lab demonstrations prove what intelligence can do, not what chemistry does on its own.

Round 2, the regress

• Q: If aliens made life on Earth, who made the aliens?
• If they say: "Maybe they evolved." Then: Evolved from what? Another cell with coded information. Where did that come from? You are in the regress.
• If they say: "Maybe they are a simulation." Then: Is the simulator embodied or not? If embodied, regress. If not, you have arrived at my conclusion.

Round 3, the escape route

• Q: You say minds cannot exist without bodies. But your own premises say bodies cannot exist without minds. Which one do you give up?
• If they give up "minds need bodies": Then you grant the conclusion is at least possible, and the evidence decides.
• If they give up "information needs minds": Then SETI, archaeology, and forensics are in trouble, because they all assume coded information implies intelligence.

Dodge table

Live-cite kit

Scripture (ASV):

• John 5:26: For as the Father hath life in himself, even so gave he to the Son also to have life in himself (John 5:26, ASV). The theological match for Stage 3: life that is underived.
• Acts 17:25: he himself giveth to all life, and breath, and all things (Acts 17:25, ASV). God as the source of all life.
• Genesis 2:7: And Jehovah God formed man of the dust of the ground, and breathed into his nostrils the breath of life; and man became a living soul (Genesis 2:7, ASV). Life from an intentional act.
• Genesis 1:1: In the beginning God created the heavens and the earth (Genesis 1:1, ASV). A finite past with a personal cause.

Scholarly:

• Stephen C. Meyer, Signature in the Cell (2009): the information argument for intelligent design in the origin of life
• Stephen C. Meyer, Return of the God Hypothesis (2021): the cumulative case from cosmology, physics, and biology
• James Tour, "We Have No Idea How Life Began" (2019): a leading synthetic chemist's assessment of the state of the field
• Douglas Axe, Undeniable (2016) and "Estimating the Prevalence of Protein Sequences Adopting Functional Enzyme Folds" (JMB, 2004): the sequence-space calculation
• Hubert Yockey, Information Theory, Evolution, and the Origin of Life (2005): information-theoretic analysis of the genetic code
• Arvind Borde, Alan Guth, and Alexander Vilenkin, "Inflationary Spacetimes Are Incomplete in Past Directions" (Physical Review Letters, 2003): the finite-past theorem

Aphorism: "Life is not an accident. It is a message. And a message always has an author."

Connection to Scripture

• John 5:26, the Father has life in Himself and gives life to the Son: the theological anchor for an underived, living source
• Acts 17:25, God gives to all life and breath: the source behind every living thing
• Genesis 2:7, God breathes life into man: the paradigm of life from a personal cause
• Genesis 1:1, the finite beginning of the heavens and the earth
• Psalm 36:9, with thee is the fountain of life: God as the living source

See also

• Arguments, the master index
• Argument from Origin of Life, the abductive case from OOL chemistry to intelligent agency (Stage 1's broader treatment)
• Biogenesis Argument, the uniformitarian case from Pasteur (life only from life)
• Lab-vs-Wild Defeater for Origin-of-Life Demonstrations, why lab results show what intelligence produces
• Minimal Cell, the floor for a self-sustaining cell
• Kalam Cosmological Argument, the finite past and the beginning of the universe
• Contingency Argument, supplies the necessity and unity the Living Source Argument does not claim
• Fine-Tuning Argument, the complementary case from physics
• Cumulative Case for Christian Theism, the larger case this argument joins