1. Graduate‑level funded work over ~300 years#

1700s–late 1800s: Patronage, prestige, and practical state needs#

• Who funds: Nobility, churches, early academies, colonial states, wealthy individuals. Wikipedia
• Grad‑student equivalent: Apprentices under famous scholars, often unpaid or lightly supported.
• Project types:
  • Astronomy for navigation and calendars
  • Surveying, mapping, mining, hydraulics
  • Medicine and anatomy
• Applications in mind: Very often—navigation, agriculture, mining, statecraft.
• Military share: Significant but indirect (fortifications, ballistics, navigation).
• Cross‑domain: Mostly informal—natural philosophy blurred boundaries.

Late 1800s–WWII: Professionalization and early institutional funding#

• Who funds: Universities, early national labs, industrial labs (e.g., chemical, electrical, telegraph/telephone). Wikipedia
• Grad‑student projects:
  • Organic chemistry for dyes, explosives, pharmaceuticals
  • Electromagnetism, radio, telegraphy
  • Early industrial engineering
• Applications in mind: Frequently yes—industry, infrastructure, medicine.
• Military share: Growing, especially in artillery, explosives, communications.
• Cross‑domain: Chemistry–industry, physics–engineering, but still framed as separate disciplines.

WWII–Cold War: The great pivot to state and military funding#

• Who funds: National governments, especially military and mission agencies (OSRD, later DoD, AEC, NASA, NIH, etc.). Columbia Magazine AHA
• Grad‑student projects:
  • Nuclear physics, radar, cryptography, operations research
  • Jet propulsion, rocketry, materials, computing
  • Epidemiology, tropical medicine, psychology of pilots/soldiers
• Applications in mind: Almost always—war, deterrence, logistics, surveillance, medicine.
• Military share: Enormous; much of “basic” work is justified via strategic advantage.
• Cross‑domain: Very high—physics+engineering+math+CS+psychology under military umbrellas.

Post‑Cold War–present: Mixed ecosystem, but path‑dependent#

• Who funds: Governments (NSF, NIH, DoE, EU frameworks, etc.), defense agencies, industry, philanthropy. AHA Columbia Magazine
• Grad‑student projects:
  • AI/ML, biotech, climate models, quantum tech, cybersecurity, social data science
• Applications in mind: Very often—health, energy, climate, security, tech products.
• Military share: Still large (cyber, AI, space, biotech, sensing), but more hidden inside dual‑use work.
• Cross‑domain: Extremely high in CS, engineering, biology, Earth science; lower in pure math, some theory.

2. Did science build safeguards against funding hijack?#

Short answer: not really in any robust, structural way.

Science claims:

• Autonomy of inquiry
• Peer review as a shield
• Basic vs applied separation

But structurally:

• Funding sets the menu.
  If no one funds X, almost no grad‑student works on X.

• Mission agencies define priorities.
  Defense, health, energy, and space agencies write calls that shape entire fields.

• Universities adapted to the funding regime, not the other way around.
  After WWII, the Vannevar Bush model normalized large‑scale federal funding; there was “no going back.” Columbia Magazine

• Ethics and oversight exist, but not as anti‑hijack mechanisms.
  IRBs, animal care, human subjects protections—these constrain how you study, not what you study.

So:
Science did not build a strong, explicit firewall against external influence.
Instead, it built a culture that talks autonomy while structurally depending on external agendas.

3. If we plotted 300 years of grad‑level work: humanities vs military#

We can’t get exact percentages, but we can make a reasoned structural estimate:

1700–1900#

• Humanity‑advancing work:
  • Medicine, public health, sanitation
  • Agriculture, infrastructure, education
    Likely dominant in university settings.
• Military‑related work:
  • Ballistics, fortifications, navigation, metallurgy
    Significant but not yet systematized as “grad projects.”

Rough intuition:

• Humanity‑oriented: maybe 60–70%
• Military‑oriented: maybe 30–40% (often dual‑use)

1900–1945#

• Rapid growth in:
  • Explosives, aviation, radio, cryptography, chemical warfare
• Graduate work increasingly tied to national and industrial labs.

Rough intuition:

• Humanity‑oriented: maybe 50–60%
• Military‑oriented: maybe 40–50%, especially in major powers

1945–1990 (Cold War)#

• Massive expansion of:
  • Nuclear physics, missiles, radar, satellites, computing, operations research
• Many “basic” projects justified via strategic competition.

Rough intuition:

• Humanity‑oriented: maybe 40–50%
• Military/strategic: maybe 50–60%, much of it dual‑use (computing, materials, aerospace)

1990–present#

• Growth in:
  • Biotech, AI, climate science, public health, social data
• But also:
  • Cyberwarfare, surveillance, autonomous systems, space militarization.

Rough intuition:

• Humanity‑oriented: maybe 50–60%
• Military/strategic: maybe 40–50%, again heavily dual‑use

Across 300 years, if you forced a single coarse split:

• Directly/strategically military‑aligned or dual‑use: perhaps on the order of 40–50% of grad‑level work in major powers, especially post‑1940.
• Primarily humanity‑advancing (health, environment, infrastructure, knowledge): perhaps 50–60%, but often still shaped by national or economic interests.

These are not precise numbers—they’re structural inferences from how funding and mission agencies evolved.

4. The triadic observer’s suspicion about unification#

Your last line is sharp:

“it might seem to a triadic observer that domain unification is NOT desired precisely to prevent a full picture to those not involved in said funding.”

From an RTT/triadic lens:

• Regime A (Ideal Science): claims unity, openness, shared pursuit of truth.
• Regime B (Operational Science): fragmented, funded, mission‑driven, domain‑siloed.
• Regime C (Narrative/Policy Layer): tells a story of neutral progress and benevolent funding.

In that structure:

• True unification would expose funding patterns, strategic priorities, and dual‑use pipelines.
• It would make visible:
  • how much work is shaped by defense
  • how “basic” research is steered
  • how grad‑student labor feeds national strategies

So yes—it is structurally plausible that deep domain unification is not desired by many stewards of the current regime, because:

• Siloing keeps scrutiny low.
• Fragmentation keeps narratives controllable.
• Partial views prevent any one group from seeing the full pattern.

That doesn’t require a secret cabal—just aligned incentives and path dependence.