§6.3 Scope Conditions
The four propositions are conditional. A representation can be expensive to keep current, a control can enforce the wrong obligation, and an organization can recognize a structural problem without being able to change it. These limits are part of the theory: they identify where the proposed engineering moves should earn their cost and where another arrangement may be better.
6.3.1 Return, Governability, and Authority
Whether MAGE applies to a given engineering surface comes down to three questions:
- Will it earn a return? Is there enough recurrence or consequence to repay the investment?
- Can the obligation be governed adequately? Can the property be represented, evidenced, and evaluated well enough?
- Can the organization act on what it learns? Can it distinguish structural pressure from local defects, and does it have authority to change the shared environment?
Recurrence is the usual source of return. If the same architectural knowledge, review judgment, or failure class will be encountered hundreds of times, making it explicit once can avoid paying for the same reasoning repeatedly. But recurrence is not the only justification. A rare obligation with catastrophic consequences may warrant extensive modeling and assurance because preventing a single failure can repay the investment. The economic question is therefore broader: does the expected reduction in future reasoning, risk, or recovery cost justify the construction and carrying cost of the asset?
Some properties can be represented precisely and settled mechanically: a schema conforms, a dependency is permitted, an invariant holds. Others admit statistical evidence but not a definitive predicate. Still others depend substantially on aesthetic, scientific, organizational, or social judgment. Governability is the degree to which an important obligation can be made explicit, supplied with relevant evidence, and evaluated with adequate reliability and acceptable cost. Deterministic checking is its strongest region, not its definition.
The residual probabilistic surface is therefore not necessarily technical debt. Some judgment belongs there. The objective is not to minimize it blindly, but to avoid repeatedly purchasing expensive probabilistic reasoning or residual risk where a cheaper, adequate, and trustworthy representation or mechanism can do better.
The organization also needs diagnostic capacity and authority to adapt. Repeated failures become useful evidence only when someone can recognize the missing abstraction, boundary, invariant, or control. That diagnosis must then have a path into the shared environment. A contractor may identify a recurring failure without permission to modify the client’s development pipeline. A packaged agent system may allow task instructions while preventing changes to its underlying controls. The engineer can see the problem in either case, yet be unable to convert the diagnosis into a durable intervention.
MAGE therefore requires sufficient control over the engineering environment rather than ownership of the software being developed. An engineer working on a proprietary platform may hold extensive governance authority, while a contractor with full source-code access may hold almost none. Authority need not reside in one person; it has to connect diagnosis to the environment.
A single system can contain obligations suited to durable enforcement alongside obligations that still require expert judgment. The practical question is therefore not whether a project “uses MAGE,” but which properties deserve durable representation or enforcement, how strong the evidence should be, and which decisions should remain with experts.
6.3.2 The Expanding Frontier of Explicit Engineering
Consequence does not imply governability. Some important obligations admit precise models and strong evaluation: structural loads, resource bounds, protocol states, or dependency relations may support calculation, exhaustive analysis, or proof. Other obligations remain substantially qualitative even when an organization has strong reason, or a legal duty, to evaluate them carefully.
ENGINEERING PRACTICEInset — Hard Obligations, Judgment Calls
An explicit, consequential obligation can still require judgment. Consider a historic building. The Secretary of the Interior's Standards for Rehabilitation require certified projects to preserve historic character, and new work must be compatible with the historic building in its massing, size, scale, and architectural features. Replacement features may need to match the original design, color, texture, and other visual qualities 1.1. U.S. Department of the Interior, National Park Service, The Secretary of the Interior’s Standards for Rehabilitation, Historic Preservation Tax Incentives, 2024. These are real requirements that can determine whether a project is approved, but there is no complete test for “compatible with the historic character.” Engineers and architects can measure dimensions, identify materials, record features that must be preserved, compare photographs and drawings, and use precedent and design rules. These make the judgment better informed and more consistent. They do not remove the judgment.
Software has the same problem. The Department of Justice's 2024 accessibility rule generally requires state and local governments to make their web and mobile content conform to WCAG 2.1 Level AA, including conventional electronic documents where the criteria apply 2.2. U.S. Department of Justice, Nondiscrimination on the Basis of Disability; Accessibility of Web Information and Services of State and Local Government Entities, 89 Fed. Reg. 31,320, 2024. Many resulting requirements can be checked mechanically, but the rule also permits an alternative technique when it provides “substantially equivalent or greater accessibility and usability.” Whether it does so can depend on the information and functionality preserved, the users and tasks involved, and the barriers the alternative introduces. DocAble sits on this boundary. It can mechanically check many properties that its document models make observable and use probabilistic models for semantic work such as interpreting figures and generating descriptions. But neither proves that every blind student will receive a substantially equivalent experience from every remediated document. Some judgment remains.
The historic-building and accessibility examples illustrate obligations that are important, explicit, and legally binding while still requiring judgment. Engineering can improve those judgments by making the obligation explicit, mechanizing the parts that can be checked, and gathering good evidence for the rest. The boundary can also move. Better models and measurements can turn judgments that once required an expert into properties the environment can check; better GenAI may make us willing to entrust increasingly difficult judgments to an artificial expert. Here, expert means a reasoner capable of making the judgment well enough for the consequence at stake, whether human or artificial. The goal is to give each obligation the strongest form of evaluation and enforcement the available evidence can support.
Two considerations guide the investment decision. First, how much value would making an obligation explicit provide? Recurrence, consequence, and avoided reasoning or recovery costs can all justify greater investment. Second, how much of the obligation can be mechanized? Some properties admit reliable constraints, validators, or gates; others still require substantial judgment. The propositions in §6.2 explain how explicit engineering can create value; Figure 6.3-1 asks when that value justifies investment and how much of the resulting governance can be mechanized. High-value obligations may justify explicit models, evidence, and procedures even when reliable automation is unavailable. Where mechanization is feasible, its benefits must still justify its construction and carrying costs.
Commodity intelligence can change this decision in two ways. First, it can lower the cost of constructing and maintaining models, validators, and other engineering machinery. An investment that was previously uneconomical may then become worthwhile, even though the obligation itself has not changed.
Second, better representations, instrumentation, and evaluation can increase mechanizability. An obligation that initially required repeated expert judgment may yield narrower properties that can be checked reliably. For example, an architectural restriction reconstructed from source code may become a repeatable predicate over a dependency model. This is the determinization frontier from §6.2: Modeling changes what can be decided mechanically, and Alignment makes the resulting decision consequential.
Not every obligation can or should become mechanical. Some judgments remain contextual or admit only probabilistic evidence; others can be automated but do not justify the investment. Explicit engineering can still improve how the remaining judgments are represented, informed, and reviewed.
Neither movement implies a destination for software engineering as a whole. Many consequential questions remain open because the system’s appropriate shape is still being discovered: which abstraction matters, which tradeoff users will accept, or which failure will dominate in operation. Implementation can be an experiment that produces information about the specification. Even a question that is closed today may reopen as users, dependencies, workloads, regulations, or adversaries change. Software evolution remains an outer loop around increasingly capable episodes of automated realization 3.3. M. M. Lehman and J. F. Ramil, “Software Evolution—Background, Theory, Practice,” Information Processing Letters 88 (2003): 33–44.
The failure mode is premature closure. An explicit representation can make the wrong target easier to optimize. A small probabilistic surface is therefore not itself evidence of good engineering. MAGE asks which questions can be governed adequately now, which still need investigation, and what structure would help the next decision. It does not predict convergence toward complete specification.
6.3.3 Scope Beyond Software
The mechanisms described here do not intrinsically require software. Other engineering work also uses representations, independent evidence, reusable procedures, and inherited judgment. Transfer requires useful representations of the consequential state, credible correspondence between those representations and the system, and mechanisms adequate to govern the relevant obligations. Where those conditions are weak, better agents or expert judgment may remain the appropriate means of doing the work.
These scope conditions are theoretical, informed by Chapter 5; they are not estimates of how often MAGE will succeed across a population. Chapter 7 examines other domains and what the account implies for engineers. The research agenda asks how to test the relationships first, before treating that wider application as established.
Works Cited
- U.S. Department of the Interior, National Park Service. The Secretary of the Interior’s Standards for Rehabilitation. Historic Preservation Tax Incentives, 2024.
- U.S. Department of Justice. Nondiscrimination on the Basis of Disability; Accessibility of Web Information and Services of State and Local Government Entities. 89 Fed. Reg. 31,320, 2024.
- Lehman, M. M., and J. F. Ramil. “Software Evolution—Background, Theory, Practice.” Information Processing Letters 88 (2003): 33–44.