Fire Origin Methodology: Challenging Negative Corpus Conclusions

The term borrows from corpus delicti. A negative corpus conclusion states a fire cause, almost always incendiary, by locating every accidental ignition source in the area of origin, eliminating each one, and then declaring that the fire must have been set by an ignition source that was never found and never physically documented. The classification is built on absence. Nothing in the scene affirmatively supports the incendiary finding. The finding survives only because the investigator ruled out the alternatives.

This inverts the logic a reliable cause determination requires. Eliminating alternatives is a legitimate step in developing a hypothesis. It is not evidence for the surviving hypothesis. A conclusion that rests entirely on what could not be found cannot be tested, because there is no artifact, residue, pattern, or measurement to test it against.

NFPA 921, Guide for Fire and Explosion Investigations, addressed this directly beginning with its 2011 edition and has carried the position forward. It treats determining a cause purely by eliminating other causes, with no evidence supporting the chosen cause, as inconsistent with the scientific method and improper. When accidental causes are eliminated and no physical evidence supports an incendiary cause, the correct classification is undetermined, not incendiary. An investigator who reaches incendiary anyway has substituted the absence of a proven accident for proof of arson.

NFPA 921 frames fire investigation as an application of the scientific method with a defined sequence: recognize the need, define the problem, collect data, analyze the data using inductive reasoning, develop a hypothesis, test the hypothesis using deductive reasoning, and select a final hypothesis. The break in a negative corpus case is at the testing step.

Data analysis is inductive. The investigator reasons from the observed scene to candidate explanations. Hypothesis testing is deductive. The investigator asks what each hypothesis predicts, then compares those predictions against the physical evidence and against all other known data, actively trying to disprove the hypothesis. A negative corpus hypothesis predicts nothing that can be checked. Its only content is that other causes were excluded, so there is no observation that could falsify it. An unfalsifiable hypothesis has not been tested. It has only been asserted.

NFPA 921 also requires the investigator to test the hypothesis against the possibility that the data are simply insufficient. That is the honest exit. When the evidence does not support any specific cause to the required level of certainty, undetermined is the scientifically defensible answer. Investigators default to incendiary under pressure, from insurers, from law enforcement, from the assumption that a fire this destructive must have been deliberate. Naming that pressure and the failure to test is the core of the challenge.

Cause cannot be determined reliably if the origin is wrong. Origin is the location where the fire began. Cause is the ignition source, the first fuel, and the circumstances that brought them together. An error in origin propagates into every downstream cause opinion, because an investigator looking in the wrong place will eliminate the wrong set of accidental sources and reach a false negative corpus.

Audit the origin basis before you engage the cause opinion at all. NFPA 921 grounds origin determination in fire patterns, witness information, electronic and video data, fire dynamics analysis, and arc mapping, the technique of plotting electrical arcing damage across a circuit to trace where the fire likely burned first and longest. Ask which of these the investigator actually used, which were documented, and whether the origin was fixed by data or assumed early and then defended.

• Was the area of origin narrowed to a defensible location, or stated as a broad region that conveniently contained a suspected set-point?
• Were ventilation effects, open doors, windows, HVAC flow, accounted for before reading burn patterns as directional evidence of origin?
• Did arc mapping support the origin, contradict it, or was it never performed on an electrically fed structure?

If the origin collapses, the negative corpus cause built on top of it collapses with it, often before you reach the methodology argument.

Many classic visual arson indicators have been shown by fire research to be products of ordinary fire dynamics, not deliberate ignition. Post-flashover and full-room-involvement conditions reproduce the same marks investigators once read as proof of accelerant.

• Irregular or pour-shaped floor patterns can be produced by radiant heat, burning plastics and synthetic furnishings melting and dripping, and ventilation-driven burning, without any ignitable liquid present.
• Low burns and burning under furniture occur when a compartment reaches flashover and the entire lower volume is involved, not only where a liquid was allegedly poured.
• Concrete spalling, crazed glass, and alligatored charring are not reliable standalone indicators of an accelerant or of rapid, hot burning and have been specifically cautioned against.

The mechanism to press is flashover. Once a compartment transitions to full involvement, the fire consumes available fuel throughout the room and generates patterns that mimic the historical arson signature. An incendiary opinion built on these patterns is challengeable unless the investigator assessed whether flashover occurred, using heat release rate, fuel load, and ventilation, and explained why the patterns are still diagnostic after that assessment. Ask directly whether the room reached flashover and how that possibility was ruled out with data rather than asserted away.

An accelerant claim should be anchored in laboratory confirmation, not visual impression. Fire debris is analyzed for ignitable liquid residue by gas chromatography-mass spectrometry under ASTM E1618, typically after passive headspace concentration under ASTM E1412. These methods identify and classify petroleum products by their chromatographic pattern. They are the affirmative physical evidence an incendiary-by-accelerant theory needs.

A negative laboratory result does not by itself disprove arson, and a competent investigator will note that ignitable liquids can burn away or evaporate. But it removes affirmative physical support for the accelerant theory and pushes the opinion back onto inference, which is exactly where negative corpus reasoning hides. Two technical points sharpen the cross-examination:

• Pyrolysis and combustion products from synthetic carpet, adhesives, and plastics can produce compounds that a careless reviewer confuses with an ignitable liquid. Confirm that comparison samples of the substrate were taken and analyzed, so background products were distinguished from an introduced accelerant.
• If the investigator asserts an accelerant despite a negative or absent E1618 result, ask what physical evidence remains for the accelerant at all. Frequently the honest answer is the burn pattern, which returns the argument to flashover and to negative corpus.

NFPA 1033, Standard for Professional Qualifications for Fire Investigator, sets the job performance requirements for the person offering the opinion. It requires the investigator to have and maintain current knowledge across a defined set of fire science topic areas, including fire chemistry, thermodynamics, fluid dynamics, fire dynamics, explosion dynamics, computer fire modeling, electricity and electrical systems, building construction, and fire analysis, among others.

This is a direct competence lever, and it interlocks with the flashover argument. An investigator cannot credibly rule out flashover, or interpret post-flashover patterns, without genuine command of fire dynamics and, where the compartment is complex, computer fire modeling. Test the witness against the specific NFPA 1033 knowledge areas rather than accepting a recitation of years on the job. Ask how they stayed current, what fire dynamics training they hold, and whether they can explain, in mechanism terms, why the patterns they relied on are still diagnostic after full-room involvement.

NFPA 921 also names expectation bias and confirmation bias as recognized threats. Probe the information sequence: who briefed the investigator, what they were told before they entered the scene, and whether the incendiary conclusion was formed before the data were collected. A conclusion that preceded the evidence is the human-factors face of negative corpus.

Negative corpus reasoning is most exposed on reliability, which is where the admissibility rules concentrate.

• Under Daubert v. Merrell Dow Pharmaceuticals, the court weighs testability and falsifiability, known or potential error rate, peer review and publication, and general acceptance. A hypothesis that cannot be tested or falsified, which is the defining feature of a negative corpus conclusion, is weak on the first and most important factor. Kumho Tire Co. v. Carmichael extends this gatekeeping to technical and experience-based testimony, so the investigator cannot escape scrutiny by framing the opinion as seasoned judgment.
• In Frye jurisdictions, Frye v. United States asks whether the methodology is generally accepted in the relevant field. NFPA 921 is widely treated as the standard of care in fire investigation, and NFPA 921 itself rejects negative corpus. That tension is usable: a method the governing guide disclaims is hard to present as generally accepted.
• Federal Rule of Evidence 702, as amended effective December 1, 2023, requires the proponent to show by a preponderance of the evidence that the opinion reflects a reliable application of the method to the facts of the case. A method that is defensible in the abstract is not enough. Negative corpus reasoning applied to a specific scene, with no evidence supporting the chosen cause, is a direct target for a motion in limine under the amended rule.

None of this guarantees exclusion, and courts vary. The point is to frame the fight on reliability and reliable application, where the methodology is weakest, rather than on the investigator's credentials alone.

Run the audit in an order that lets each finding set up the next.

1. Assemble the record. Obtain the full report, the investigator's field notes, all scene and laboratory photographs, and every lab report. Confirm whether any ignitable liquid was confirmed under ASTM E1618 or only asserted.
2. Split origin from cause. Force the witness to state the origin opinion and its data separately from the cause opinion. Establish whether origin was fixed before cause, and on what evidence.
3. Test the eliminations. For each accidental source said to be excluded, ask whether it was physically located, examined, and documented, or merely asserted as eliminated. An elimination without documentation is not data.
4. Force the flashover question. Ask whether the compartment reached flashover, how that was assessed, and why the relied-upon patterns remain diagnostic afterward.
5. Name the missing evidence. Ask what affirmatively supports the incendiary cause. If the answer reduces to elimination plus burn patterns, you have surfaced the negative corpus on the record.
6. Set up the motion. Tie the untested, unfalsifiable hypothesis to Daubert and to the FRE 702 reliable-application requirement, and preserve the NFPA 921 and NFPA 1033 deviations as the standard-of-care backdrop.

This is procurement and litigation-support analysis, not legal advice, and no methodology critique guarantees exclusion or a given outcome. Retain a qualified fire investigation expert to apply NFPA 921 to the specific scene.