When Protocols Endure
A Considered Response to "The Hard Authority of the Protocol" re:The Uneven Fate of Codified Judgment
Dear colleague,
Your essay opens with a precise and generous diagnosis. The image of coordination failures arriving “with receipts” captures a contemporary unease that many organizations recognize immediately: the unsettling experience of having complied fully with procedure while feeling that competence has thinned out somewhere along the way. The essay’s insistence that this condition is a quiet scandal, one that passes as modernization rather than a change in what coordination is, does important work. It names a shift that otherwise remains tacit. The move to treat protocols as legible artifacts and practice as a living, time-bound accomplishment grounds the argument in a vocabulary that allows the problem to be seen rather than merely felt.
That framing gains further traction through your use of the Summer of Protocols definition of protocols as engineered arguments designed to stabilize tensions. By pairing that definition with Wenger’s account of communities of practice, you establish an asymmetry that feels both intuitive and analytically productive. Practice appears as the site where judgment is developed and sustained through participation, while protocols appear as condensations of that judgment, packaged for travel and enforcement. The appeal of Weick’s process–product distinction in this context is clear. Freezing parts of practice into portable artifacts buys reach and continuity, and it also buys context loss. The essay is strongest where it resists treating that loss as a moral failure and instead treats it as the price of abstraction.
The point at which I want to slow the argument is where this asymmetry is allowed to harden into a general tendency. The essay’s core claim suggests that once protocols become executable and operationally dominant, decay follows when they are detached from the communities of practice that generated them. This claim is compelling in the Git-centered case you develop, and the Git case is developed with care. It is also where the argument risks overextending itself. The risk does not come from exaggeration or polemic, but from a subtle collapse of cases that differ in their relationship to judgment, closure, and maintenance.
Git functions in your essay as an anchor example of a protocol that emerged from a managerial practice and was later subsumed into software, acquiring a form of authority that outstrips the practice that once sustained it. That reading is persuasive. Git encodes judgments about authorship, sequencing, conflict resolution, and responsibility. Commits and branches stabilize tensions between autonomy and coordination, while merge conflicts externalize disagreement into a constrained procedural space. As Git becomes embedded in toolchains, identity systems, compliance regimes, and dashboards, the protocol begins to define what counts as real work. Judgment migrates from action into configuration and exception handling. The protocol becomes an environment.
In that environment, the dynamics you describe become visible. Reviews turn into rubber stamps. Commit histories lose interpretability. Work that resists protocol legibility is quietly discouraged. Enforcement scales faster than care. Decay, understood as a widening gap between enforced form and competent action, becomes plausible and, in many settings, observable. The essay’s account of this drift is convincing because Git remains deeply dependent on situated interpretation at runtime. A merge conflict still requires someone to decide what coherence means in context. A passing test suite still requires someone to decide whether the system is sound. The protocol coordinates behavior, and practice supplies meaning.
The difficulty arises when the essay’s diagnosis is allowed to stand in for a more general theory of protocols under operational dominance. In several places, Git’s dependence on runtime judgment is treated as representative rather than as diagnostic of a particular regime. Protocol dependence on practice becomes universal, and dominance becomes tightly bound to decay. This move is understandable, given the empirical richness of the Git case, and it is also where a further distinction would sharpen the theory rather than weaken it.
The distinction I have in mind concerns the degree to which a protocol exhausts judgment upstream rather than deferring it to runtime. Some protocols remain porous at the point of execution, relying on situated interpretation to complete their work. Others aim to close key ambiguities in advance through strong invariants and institutional scaffolding. The consequences of operational dominance differ markedly between these cases.
Double-entry bookkeeping provides a useful counter-case, not because it contradicts your argument, but because it refines its scope. Double-entry emerged from managerial needs that resemble those you attribute to Git: accountability, control at a distance, reconciliation, and error detection. It stabilized recurring tensions between trust and verification, autonomy and oversight, speed and accuracy. Its core invariant is deceptively simple: every entry has equal debits and credits, and the books must balance. Beneath that lies the accounting equation, which constrains what can count as a valid representation of economic activity.
This invariant closes a large class of ambiguities upstream. An entry either balances or it does not. Errors surface mechanically through imbalance. Responsibility becomes traceable through the structure of the ledger. The protocol was disseminated through codification in the late fifteenth century, often associated with Pacioli, and it persisted because it solved a coordination problem that recurred across contexts. More importantly, it was embedded in an institutional ecology that relocated maintenance away from local practice without eliminating it. Education, professional standards, audit regimes, and later regulatory frameworks formed a scaffolding that sustained the protocol’s coherence over centuries.
When double-entry became subsumed into software, the pattern you describe for Git also appeared, with a crucial difference. General ledgers and ERP systems embed the accounting protocol so that operational events generate postings automatically. The ledger defines legibility. Transactions that do not post do not exist organizationally. Authority shifts toward system enforcement, and the protocol becomes an environment. Yet widespread decay in the sense you describe does not follow in the same way. The books continue to balance. Errors continue to surface. Coordination persists across vast scales and across communities that share little in the way of lived practice.
This persistence does not imply that practice has disappeared. Judgment remains in classification, valuation, estimation, and boundary cases. Decisions about revenue recognition, asset impairment, and provisions require interpretation and are sites of contestation. What differs is where that judgment is exercised and how it is maintained. In double-entry, much of the maintenance work is relocated upstream into standards-setting bodies, audit practices, and professional education. Repair occurs through revisions to standards and through institutionalized review rather than through ad hoc runtime interpretation by local communities. The protocol remains ontologically dependent on practice, though the practice is distributed and institutionalized rather than localized and continuous.
Seen through this lens, the dynamics you describe in the Git case exemplify a practice-dependent protocol regime. Git requires ongoing situated interpretation to function competently, and when operational dominance expands faster than the community’s capacity to supply that interpretation and repair, drift emerges. Double-entry exemplifies a more practice-exhaustive regime. Key ambiguities are closed through formal invariants, and maintenance is routed through institutional channels that can persist beyond local communities of practice.
This distinction matters because it reframes the relationship between dominance and decay. Dominance does not, by itself, produce decay. Decay becomes predictable when operational dominance expands in settings where runtime judgment remains essential and where the structures that sustain that judgment are weakened or displaced. Where judgment has been largely exhausted upstream and supported by durable scaffolding, operational dominance can coexist with long-term coherence.
Your essay already contains the resources for this refinement. The emphasis on protocols as engineered arguments managing tensions invites the question of how much of the argument is resolved in advance and how much is deferred to execution. The discussion of maintenance as a meta-protocol points toward different maintenance topologies rather than a single failure mode. Wenger’s insistence that communities of practice produce living knowledge, coupled with his warning about treating repositories as substitutes for participation, aligns closely with the Git case and less neatly with double-entry, where repositories and standards play a constitutive role. Weick’s caution about frozen products traveling farther than the processes that generated them remains central, and it suggests attending to where and how the costs of freezing are paid.
Re-reading Git through this revised lens clarifies why your decay diagnosis fits so well. Git’s protocol leaves much of the work of sensemaking to runtime practice. When the protocol becomes the primary object of enforcement and measurement, the social processes that supply care and repair are easily underprovided. Double-entry, by contrast, shifts a significant portion of that care upstream. The protocol’s success depends less on continuous local negotiation and more on slow-moving institutional revision. Each regime extracts context, and each pays for that extraction differently.
Framed this way, the disagreement becomes productive rather than oppositional. Protocols extend practice by increasing capacity, reach, durability, and auditability. In some domains, they also replace large portions of runtime practice by closing ambiguities in advance. The analytic task is to specify where judgment belongs, where it can be safely eliminated, and where its elimination changes the nature of the work being coordinated.
Your essay makes a persuasive case that many contemporary organizational failures arise from treating practice-dependent protocols as though they were practice-exhaustive, allowing enforcement to scale while care does not. The addition of a two-regime distinction sharpens that diagnosis without blunting its force. It suggests that the problem is less the hard authority of protocols in general and more the misalignment between the kind of authority a protocol exercises and the kind of judgment the domain continues to require.
Restoring proportion, in this sense, does not mean choosing between practice and protocol. It means attending to the different ways protocols relate to practice over time. Freezing buys reach. Closure buys durability. Each extracts context. The costs appear either in local repair or in institutional redesign. When those costs are misallocated, coordination can persist while competence thins out. When they are aligned, protocols can travel far without leaving emptiness behind.