Virology Methodology
The standard methodology for establishing viral pathogenesis follows a series of steps, each of which takes the output of the previous step as its input. This entry traces the dependency chain, identifies where inference substitutes for demonstration at each link, and states what verification would look like at each gap.
The structure of this entry is independent of which theory of disease is ultimately correct. It maps where observational proof exists and where it is absent, so that any model — germ theory, terrain theory, or otherwise — can be evaluated against the actual evidentiary record rather than against institutional confidence.
How Claims in This Entry Are Weighted
Claims here are labelled by inspection status — whether the primary material was read — and not by publication venue. Any reader applying a different standard should make that standard explicit.
Publication venue is a weak criterion for this subject specifically. Peer Review sets out why: the review system functions as a credentialing gate whose reviewers are drawn from the paradigm under examination, and Input Constraint sets out why work challenging a foundational model is filtered upstream of review, at funding and institutional affiliation. Dismissing such work because it was not published in a reviewed venue therefore assumes what is in question. The absence is the predicted output of the filter, not independent evidence about the work.
That argument is reciprocal, and it costs this entry something. It does not convert unreviewed work into established fact. The control experiments cited here from Lanka and Andrews are marked reported, because their protocols and data have not been inspected for this entry — the same label that would be applied to an unread paper in a reviewed journal. Venue neither rescues nor condemns a claim.
What does count against a claim, in either direction:
- The protocol is not documented sufficiently for a third party to repeat it
- The materials, conditions, or negative controls are unstated
- Results that would bear on the conclusion are withheld
- The data are contradicted by an identifiable record
Openly published protocols and data satisfy the first three directly, and are inspectable without an intermediary vouching for them. Fully documented open work is, on this standard, easier to check than a reviewed paper whose underlying data are not released.
The Dependency Chain
Step 1: Clinical observation
| Observed | A person presents with specific symptoms. |
|---|---|
| Inferred | The symptoms are caused by a specific transmissible pathogenic agent. |
| Gap | Symptoms can have multiple causes. The same symptom cluster may appear in different contexts and from different etiologies. The grouping of symptoms into a named "disease" is itself a classification decision, not an observation. |
| What would close the gap | A demonstrated causal mechanism linking a specific agent to the specific symptom cluster, distinguishing it from other causes of the same symptoms. |
Step 2: Sample collection
| Observed | Biological material (mucus, blood, tissue) is collected from the sick person. |
|---|---|
| Inferred | If a pathogenic agent exists, it is present in this sample. |
| Gap | The sample is a complex mixture: host cells, bacteria, cellular debris, proteins, exosomes, genetic material from the host and from microbiome organisms. The target agent has not been identified or separated at this stage. |
| What would close the gap | This step is preparatory. The gap is acceptable here provided subsequent steps achieve separation and identification. |
Step 3: Cell culture and cytopathic effect (CPE)
| Observed | The sample is added to a cell culture (typically Vero cells, HEp-2, or similar). Antibiotics (gentamicin, amphotericin B, or similar) and nutrients (often fetal bovine serum) are added. The culture is starved or stressed. Cells deteriorate and die — this is declared as cytopathic effect. |
|---|---|
| Inferred | The CPE is caused by a virus in the sample replicating within and destroying the cells. |
| Gap — critical | Cell deterioration in culture is not by itself attributable to an inoculum, and the foundational literature says so explicitly.
Enders and Peebles (1954), the paper that established the tissue-culture method for measles, reports encountering a second agent obtained from an uninoculated culture of monkey kidney cells whose cytopathic changes, in unstained preparations, "could not be distinguished with confidence" from the agents isolated from measles patients. The authors state that the possibility of encountering such agents "should be constantly kept in mind." The same paper reports two means by which that second agent was distinguished, and both must be stated:
Three qualifications bear on how much those discriminators carry:
Correction of a common misstatement: the uninoculated controls illustrated in the 1954 paper show normal cell outgrowth, in both fresh and stained preparations, and are presented as controls for the inoculated cultures. The contaminating agent was a separate finding, not the behaviour of the routine control. Claims that Enders' controls themselves reproduced the measles cytopathic effect misdescribe the paper. Independently of Enders, control experiments by Stefan Lanka and documented open-source replications by Jamie Andrews are reported to produce comparable cell deterioration under nutrient restriction and antibiotic exposure without a patient sample. These are reported rather than observed here because the primary protocols and data have not been inspected for this entry. Their publication venue is not the basis for that label; inspection status is. |
| What would close the gap | A control run under conditions identical to the inoculated culture in every respect except the patient sample, showing no comparable deterioration — with the discrimination performed at the same post-inoculation interval in both arms, since the reported nuclear discriminators are time-dependent. |
Step 4: Declaration of "isolation"
| Observed | Material from the culture showing CPE is collected (typically the supernatant — the liquid above the cells). |
|---|---|
| Inferred | This material contains the virus. The virus has been "isolated." |
| Gap — critical | "Isolation" in virology does not mean what the word means in chemistry, microbiology, or common usage. It does not mean the particle has been physically separated from all other material. It means the sample has been passaged through cell culture and CPE has been observed. The supernatant collected contains: remnants of host cells, culture medium, antibiotics, fetal bovine serum, exosomes, cellular debris, and potentially the target particle — unseparated.
Physical purification of particles is an established and available technique — density-gradient ultracentrifugation in sucrose, caesium chloride or iodixanol, filtration, and chromatography. This entry does not claim otherwise, and an entry that did would be wrong. The question is narrower and checkable paper by paper: was purification performed before the material in question was characterised or sequenced? Where it was not, the characterisation inherits whatever the culture or clinical specimen contained. A documented instance: Wu et al. (2020), the paper describing the first published genome of the agent associated with the Wuhan respiratory disease outbreak, extracted total RNA directly from a bronchoalveolar lavage fluid specimen using a commercial column kit, and assembled the genome by meta-transcriptomic sequencing of that extract. No particle purification step preceded sequencing. The resulting genome was assembled computationally and checked by alignment against existing coronavirus sequences. This is a statement about that paper, not about virology as a whole. The productive form of this objection is a survey: for a given agent, identify which foundational characterisation papers purified particles before sequencing and which did not. That survey does not exist in this entry and is not asserted here. |
| What would close the gap | For any specific agent: a characterisation performed on particles physically separated from culture and host material, with purity independently confirmed, and the sequence derived from that purified preparation rather than from the mixed specimen. |
Step 5: Electron microscopy
| Observed | Particles are visible in electron micrographs of culture material. |
|---|---|
| Inferred | These particles are the virus. |
| Gap | The EM images are typically of culture supernatant or thin-sectioned culture material — not of purified particles. The culture contains cellular debris, exosomes, and other particles of similar size (30–150 nm) and morphology. Without prior purification, the identification of specific particles as "the virus" is an assumption based on morphological resemblance to what the model predicts the virus should look like. Immunogold labelling can distinguish particles by antibody binding, but the antibody's specificity is itself established through the same chain. |
| What would close the gap | EM of purified, characterised particles confirmed as distinct from normal cellular debris and exosomes — and confirmed as the same entity whose genetic content was sequenced (see Step 6). |
Step 6: Genome sequencing
| Observed | Genetic material is extracted from the culture material and sequenced. For RNA viruses, short fragments (reads) are computationally assembled into a genome. |
|---|---|
| Inferred | The assembled sequence is the genome of the virus. |
| Gap | The genetic material is extracted from the same impure culture material containing host cell DNA/RNA, bacterial genetic material, and other biological content. Computational assembly, particularly for RNA viruses, uses algorithms that construct a genome from short fragments, often guided by reference sequences for alignment. The reference sequences were themselves derived from the same methodology applied to earlier cultures. A genome assembled from mixed material and aligned against a computationally derived reference is a model of a model. |
| What would close the gap | Sequencing of genetic material extracted from purified, physically separated particles whose identity has been independently confirmed. The sequence must be derived from the particle, not from the culture mixture. De novo assembly without reference-guided alignment would provide stronger evidence of an independent genome. |
Step 7: PCR primer design
| Observed | Short sequences (primers) are designed to match specific regions of the genome from Step 6. |
|---|---|
| Inferred | These primers will specifically detect the virus in patient samples. |
| Gap | The primers detect a specific short nucleotide sequence. That target sequence was identified through Steps 3–6. If any step in that chain contains an unverified inference, the primers may detect a sequence that exists in nature but has not been demonstrated to belong to a pathogenic virus. PCR amplifies — it does not identify the biological origin or function of what it amplifies. |
| What would close the gap | Validation that the PCR target sequence is exclusively present in the purified, characterised pathogenic particle, and absent from healthy tissue, normal cellular material, and microbiome organisms. This requires the purification from Step 4 to have been completed. |
Step 8: Diagnostic application
| Observed | A PCR test or antigen test returns a positive result on a patient sample. |
|---|---|
| Inferred | The patient is infected with the pathogen and may be a "case" of the disease. |
| Gap | A positive PCR result means the target sequence was present above the cycle threshold (Ct). It does not independently confirm: (a) that the sequence belongs to a virus, (b) that the virus is pathogenic, (c) that the patient is sick because of it, or (d) that the patient is infectious. The cycle threshold is a methodological choice — higher Ct values detect smaller quantities of the sequence, increasing sensitivity but also increasing detection of fragments that may have no clinical significance. For antigen tests, the antigens targeted are proteins identified from the same methodology chain. |
| What would close the gap | Clinical validation demonstrating that a positive test result reliably correlates with (a) the presence of a characterised pathogenic particle, (b) active disease in the patient, and (c) capacity to transmit. This requires the upstream chain to be independently validated first. |
Step 9: Case counting and model validation
| Observed | Positive test results are counted as "confirmed cases." |
|---|---|
| Inferred | The case count validates the model of a circulating pathogen causing disease. |
| Gap — structural | If the test detects a sequence whose identity as a pathogen was established through the chain above, then counting positive tests is counting the methodology's own outputs. The count cannot independently validate the model because the count is produced by the model's own tools. This is the self-referential closure of the chain. |
| What would close the gap | Case definitions based on clinical presentation (symptoms, severity, outcome) correlated independently with test results — and the test results themselves validated against the upstream chain. |
The Self-Referential Loop
The complete chain forms a closed loop:
- CPE in culture → declared as evidence of virus
- Culture material → particles assumed to be virus
- Particles → sequenced to produce a genome
- Genome → used to design diagnostic tools
- Diagnostic tools → produce positive results
- Positive results → counted as cases
- Case counts → treated as validation that the virus exists and causes disease
- Validation → justifies continued use of CPE methodology as the standard
Within the diagnostic chain, no step is checked against an external reference. Each step's output is the next step's unexamined input, and the case count cannot validate the model that produced it.
One method does break this closure, and it must be addressed rather than absorbed into the loop: reverse genetics reverses the direction of inference. In the chain above, the sequence is a descriptive output derived from culture. In an infectious clone, the sequence is a causal input from which a replicating agent is recovered. That is a genuine external check on one specific link. It does not close the chain's other gaps, and what it demonstrates is narrower than it is usually taken to show.
Reverse Genetics
An infectious clone is a DNA copy of a viral genome that can be introduced into cells, from which a replicating agent is then recovered. The technique inverts the direction of inference in the dependency chain: instead of deriving a sequence from a culture, it begins with the sequence and produces the agent.
This is the strongest available objection to the self-referential reading, and the two cases below are reported in detail because the argument should be met at its strongest rather than at its weakest.
Poliovirus (1981)
Racaniello and Baltimore inserted a complete cDNA copy of the poliovirus genome into the plasmid pBR322. Mammalian cells transfected with the plasmid produced infectious poliovirus. A construct lacking the first 115 bases produced none.
The negative control is the significant part. Altering the sequence while holding every other condition constant produced no agent. That is a method producing a negative result, keyed to the sequence.
Zika virus (2020)
A bacterial artificial chromosome carrying the sequence of ZIKV strain Paraiba/2015, under a cytomegalovirus promoter, was delivered into mice. The relevant details:
| Element | As performed |
|---|---|
| Host | IFNAR−/− A129 mice, 4–5 weeks old, six per group (three male, three female) |
| Routes | Subcutaneous in the footpad, intramuscular in the quadriceps, intraperitoneal in the peritoneal cavity |
| Inoculum (first design) | 1 × 10⁶ Vero cells per mouse, transfected with the plasmid |
| Inoculum (second design) | Plasmid complexed with Lipofectamine 2000, up to 100 µg per mouse in 200 µl Opti-MEM |
| Outcome | Rapid weight loss from about 6 days post-inoculation, severe paralysis, death by 8–10 days |
| Re-isolation | Agent recovered from sera of inoculated mice retained virulence in a second round of infection in new mice |
The control arms
The obvious objection is that injecting foreign material into an animal provokes an inflammatory response, so illness following injection does not identify the sequence as its cause. The study's controls address this objection directly, and the results do not support it:
- Empty-plasmid controls. Mice inoculated with Vero cells transfected with empty plasmid, and mice inoculated with empty plasmid directly, showed no clinical signs, no weight loss, no death, and no recoverable agent at any timepoint. The injection, the cells, the transfection reagent, the mouse strain and the routes were all held constant; only the sequence was removed.
- The transfection reagent was present in the control arms. Cationic lipid reagents are themselves inflammatory. That confound was therefore included in the negative controls, and produced no disease.
- Dose-dependence. Doses of 100, 50, 25 and 12.5 µg were uniformly lethal; 6.25 µg largely was not. Injection trauma does not scale with the quantity of plasmid delivered.
- Route-dependence runs counter to the objection. Intraperitoneal delivery was the most lethal and intramuscular the least. A generic reaction to injected non-self material would not be mildest when injected into muscle.
- Second passage. Serum from affected mice caused disease in new mice. That material carries neither plasmid nor transfection reagent.
Taken together, these controls isolate the sequence as the operative variable. The objection "they injected it" is true as a description and insufficient as an explanation, because the empty-plasmid arm is that objection with the sequence removed, and it produced nothing.
What this establishes, and what it does not
The distinction matters, because the first claim is routinely allowed to stand in for the second:
| Claim | Status |
|---|---|
| This sequence is sufficient to cause disease in an interferon-receptor-deficient mouse when delivered by injection | Demonstrated, with controls |
| A purified particle transmits disease between immunocompetent hosts by the natural route | Not addressed by this experiment |
The departures from the specification in The Missing Experiment are nameable, and none of them is incidental:
- The host was engineered to be susceptible. A129 mice lack the type-I interferon receptor. The model is used because the agent does not reliably produce disease in immunocompetent mice. An experiment in a host built to be unable to mount a standard response cannot establish what happens in one that can.
- No natural route was used. The agent is described as primarily mosquito-borne, which deposits material intradermally during blood-feeding. Footpad, quadriceps and peritoneal injection reproduce none of this. The paper describes its routes as those "typically used" for the agent — an indication that the standard is convention, not natural exposure.
- The starting material is a DNA sequence, not a purified particle. Steps 4 and 5 of the chain — physical purification and characterisation of a particle — are not performed and not required by this method. The technique bypasses the purification gap rather than closing it.
- Transmission was not tested. Re-isolation and re-inoculation by injection is not transmission between hosts.
Consequence for this entry
Reverse genetics defeats the strong form of the self-referential claim. It does not satisfy the specification in The Missing Experiment, and it leaves the purification and natural-route gaps untouched.
What would close the remaining gap: the same sequence-level controls, applied to a purified particle introduced to an immunocompetent host by the claimed natural route, with transmission to further hosts measured rather than inferred.
The Missing Experiment
The experiment that would close the causal chain has the following form. Nothing in it is unusual; it is ordinary experimental design applied to a question where it is frequently not applied.
- Physically purify the particle from all other material
- Characterise the purified particle independently
- Expose a healthy, immunocompetent subject through the claimed natural infection pathway — external exposure, not injection, since injection bypasses the barriers the claimed route must cross
- Run a process-matched control: subjects exposed to the complete preparation process and all of its reagents, with the particle removed and nothing else altered
- Allocate subjects prospectively and assess outcomes blinded, with protocol, endpoints and case definitions fixed before exposure begins
- Compare disease rates between the arms, rather than reporting the exposed arm alone
- Measure onward transmission to further subjects, and re-isolate
Items 4 and 5 carry most of the weight, and are the ones most often absent. A control consisting of the full preparation process minus the particle is what distinguishes the particle from everything else delivered alongside it. Prospective allocation and blinded assessment are what prevent the expected result from shaping which outcomes are recorded as the disease.
Reading the possible outcomes. The design is only useful if each result means something definite in advance:
| Result | What follows |
|---|---|
| Disease in the exposed arm at a rate materially above the process-matched control | The causal claim is supported for that route and that host |
| Comparable illness in both arms | The causal claim is falsified. The preparation process, not the particle, accounts for the outcome |
| No illness in either arm | The causal claim is unproven, not refuted |
The third row is stated against this entry's own interest, and deliberately. A null result is consistent both with there being no causal agent and with there being one whose attack rate is low in a group whose susceptibility was never established. Treating a null as disproof would be the same error, in the opposite direction, as treating an unpurified culture as proof. Separating those possibilities requires susceptibility to be established independently, which is a further experiment again.
Stating that this has never been attempted would be false, and this entry does not state it. Attempts exist. Each departs from the specification in ways that are nameable, and naming them is more useful than an absolute claim:
| Attempt | What was done | Departures from the specification |
|---|---|---|
| Enders and Peebles, 1954 | Two litters of suckling mice inoculated intraperitoneally and intracerebrally with infective tissue culture fluid | The animals remained well through 21 days of observation. No disease was produced. Routes were injection, not natural exposure. Material was culture fluid, not a purified particle. |
| Human challenge study, 2022 | 36 volunteers aged 18–29 inoculated intranasally with a standardised dose of a wild-type agent, then housed in a quarantine unit with continuous monitoring | Open-label and non-randomised, with no placebo-inoculated control arm. The inoculum was culture-derived, not a purified particle. Intranasal instillation of a measured dose is not the claimed natural route; the authors present this standardisation as a deliberate contrast with field and household-contact studies, in which exposure conditions are unknown. |
| Reverse-genetics studies | See Reverse Genetics | Sequence rather than purified particle as starting material; injection routes; in the Zika case an interferon-receptor-deficient host. |
The authors of the foundational paper said the same thing. Enders and Peebles close by stating that two experiments "essential in the establishment of this relationship remain to be carried out": production of the disease in monkey and in man using tissue-culture material passaged enough times to eliminate any agent carried in the original inoculum, followed by recovery of the agent from the experimentally produced disease. Their stated conclusion is that the findings "support the presumption" that these agents represent the viral species responsible for measles. The paper also urges that "great caution" be exercised in interpreting claims that the agent has been propagated in other hosts or systems.
This entry's claim is therefore not that the experiment is unimaginable or untried. It is that the specified experiment — purified particle, natural route, immunocompetent host, transmission measured, placebo-controlled — is not what the foundational literature rests on, and that the authors of that literature said as much at the time.
Addressing Common Rhetorical Responses
"You're moving the goalposts"
This accusation assumes each method (CPE, sequencing, PCR) is an independent line of evidence. If three independent paths converge on the same conclusion, then rejecting all three does look like motivated skepticism.
But the methods are not independent. They are a dependency chain — each step takes the previous step's unverified output as its input. Challenging each link is not attacking three separate things. It is tracing one chain and showing that no link is independently verified.
Moving goalposts means changing one's criteria after they have been met. Tracing a dependency chain means showing the criteria were never met at the foundation.
"Your position is unfalsifiable"
The position outlined here states explicit falsification criteria:
- Purify the particle
- Characterise it independently
- Reproduce the disease through the natural pathway with adequate controls
If this experiment were conducted and the disease reproduced reliably, the position falls.
The claim that the method can never produce a negative result is too strong, and is replaced here by the narrower claim it can actually support.
Negative results are producible, and the 1954 paper itself produced several. Convalescent sera from two measles patients neutralised the cytopathogenic effect of the agent isolated from a third, with neutralising titres rising between acute and convalescent samples. Control antigens taken from uninoculated cultures failed to fix complement with any serum tested. A contaminating agent was identified and excluded by two independent criteria. A filtrate passed through a sintered glass filter shown to retain Serratia marcescens retained its cytopathic activity, and that activity was destroyed by heating to 65 °C.
What this means for the entry's argument: the method can discriminate between agents. It does not follow that it can establish pathogenicity. Neutralisation, complement fixation and filterability are all assays of behaviour in culture. They answer the question "is this the same agent?" They do not answer the question "does this agent cause this disease in a host?" The unfalsifiability concern is therefore stated against the second question, where it survives, and withdrawn from the first, where it does not.
"Modern sequencing has moved beyond culture"
Metagenomic sequencing can detect sequences directly from patient samples without culture. However:
- The sequences detected are identified by alignment against reference genomes
- The reference genomes were assembled from culture-derived material using the methodology described above
- A sequence match against a reference is only as reliable as the reference
- The reference traces back to the same chain
Direct sequencing is a technical advance in detection, but it does not provide an independent check on the identity or pathogenicity of what is being detected if the reference it matches against is itself derived from the unverified chain.
"Koch's postulates are outdated"
Koch's postulates have been declared inapplicable to viruses because viruses require host cells to replicate and because some infections are asymptomatic. These are methodological complications, not exemptions from the requirement to demonstrate causation. The core logic — isolate the agent, reproduce the disease, re-isolate the agent — is not a historical curiosity. It is the minimum standard for establishing that a specific agent causes a specific disease.
Relaxing the standard does not strengthen the evidence. It weakens the requirement for evidence.
Epistemic Status
This entry maps a methodological dependency chain and identifies where inference substitutes for demonstration. It does not assert that no virus has ever caused disease. It asserts that the standard methodology, as practised, does not close the causal chain at its foundation, and that each subsequent layer inherits this gap.
The Enders and Peebles (1954) material is observed: the full paper was read, including its figures, control descriptions, neutralisation and complement-fixation results, filtration and thermal-stability data, and its closing statement of what remained to be done. The Wu et al. (2020) sequencing method and the 2022 challenge-study design are observed from the published methods. The Zika reverse-genetics study is observed from its full text. The Racaniello and Baltimore (1981) result is reported from an identified source rather than from the paper itself.
Claims about Lanka's and Andrews' control experiments are reported and not independently inspected here.
Whether the patterns described generalise beyond the specific papers cited is unresolved. This entry surveys named studies; it does not survey the field, and no claim about "all" or "no" virology studies should be read into it.
If the foundation is sound, these gaps should be closable through the experiments described. If the foundation is not sound, no amount of downstream sophistication (better sequencing, faster PCR, larger case counts) can compensate, because each downstream method takes the foundation's output as its unexamined input.
The question is not whether the model is believed. The question is whether the model is demonstrated.
References
- Enders JF, Peebles TC. "Propagation in Tissue Cultures of Cytopathogenic Agents from Patients with Measles." Proceedings of the Society for Experimental Biology and Medicine 86:277–286 (1954).
- Racaniello VR, Baltimore D. "Cloned Poliovirus Complementary DNA Is Infectious in Mammalian Cells." Science 214:916–919 (1981).
- Wu F, Zhao S, Yu B, et al. "A new coronavirus associated with human respiratory disease in China." Nature 579:265–269 (2020).
- Killingley B, Mann AJ, Kalinova M, et al. "Safety, tolerability and viral kinetics during SARS-CoV-2 human challenge in young adults." Nature Medicine 28:1031–1041 (2022).
- "In vivo rescue of recombinant Zika virus from an infectious cDNA clone and its implications in vaccine development." Scientific Reports 10 (2020). doi:10.1038/s41598-020-57545-2