Bottom-Up Doctrine Generation
Question. Does the civilian network actually surface doctrine that DRDC's own research cohorts cannot generate, or does it just reinvent what the institution already has?
Analogy. It works like the fixes soldiers rig to their own kit in the field: a soldier solves a real problem in a way the central design bureau never drew, and sometimes the bureau studies it and issues it as standard kit, while other times it turns out to be a rougher version of gear the armoury already stocks. The study asks whether the civilian network working real problems surfaces techniques the institution's own research cohorts did not and could not have produced, or whether it just reinvents doctrine DRDC already holds.
What's at stake. This is the larger half of the proposal's case. The defence story justifies the funding, but the bet that pays it back is bottom-up innovation: the claim that a civilian network working real problems generates genuinely new ideas that top-down, managed research cannot, and that this is the national multiplier that grows the economy and the tax base the defence is funded from. The whole fund-the-seed-once argument rests on the network actually generating something the institution could not have generated itself. The proposal asserts this, Lawrence's Rule run in reverse, with civilian field doctrine flowing to institutional use, and no study tests it. If the network just reinvents what DRDC's own programs already produce, the innovation multiplier is rhetorical and the proposal is a resilience instrument only.
The two answers it decides between. Either naturalistic real-stakes practice explores a space managed cohorts cannot reach, so the network surfaces field-tested techniques that are both genuinely novel and that the institution's internal programs did not and could not produce (Lawrence's Rule); or the network reproduces institutional doctrine, or its novel-looking techniques are folk variants without rigour, so there is no institution-could-not-make-this output (replication). A blind comparison of the two corpora tells them apart.
What a null result would mean. If a blind expert panel cannot find techniques in the network's corpus that are both genuinely novel and beyond what DRDC's own cohorts produced over the same period, the bottom-up-innovation and Lawrence's-Rule-reversed claims need revision: the network is a training and resilience instrument, not an innovation source. That is a finding about the innovation theory the proposal's larger half rests on, not a sign the platform was built wrong.
Why this matters to defence. The proposal asks DRDC to treat the same spend as buying both cognitive-warfare resilience and a national innovation multiplier that eventually funds the defence (DRDC Objective 4 and self-funding sustainability; DRDC Objective 6 and the evolution-of-science-and-technology focus area). The multiplier half depends entirely on the network producing what the institution cannot. The result changes a concrete decision: whether to fund and harvest the civilian network as a doctrine and innovation source, or to price the proposal as a resilience instrument and keep innovation inside internal R&D.
How we would run it. Over the same period, collect two corpora: the technique catalogue the network surfaces, the doctrine that gets contributed and actually adopted by other operators, and the output of DRDC's own managed research cohorts working comparable problems. Strip both of any mark of where they came from and give them to an expert panel who do not know which corpus is which. Have the panel rate each technique for genuine novelty, for whether the institution's own programs did or realistically could have produced it, and for whether DRDC would adopt it. The discriminating measure is whether the network corpus contains techniques the panel judges novel-and-institution-could-not-have-made that the managed-cohort corpus lacks. The blinding is what keeps "the network is more creative" from being assumed rather than shown. The only-us part is structural: the civilian network working real problems under naturalistic stress is, by definition, the thing a managed institutional cohort is not.
Earliest start. Stage 16: the study needs a mature compendium of network-adopted doctrine and a comparable managed-cohort corpus to compare it against.