Three months ago, I had a stroke. Minor, as strokes go—my cognition is intact, my body largely functional. But something has changed in the wiring, and what I am observing in my own recovery has reframed how I think about institutional fragility and why complex systems fail where we least expect them to.
The Word I Cannot Find
Here is what a minor stroke actually feels like from the inside.
My brain knows the word. I can feel its shape, its weight, its approximate location in memory. But the automatic retrieval pathway—the one that ordinarily delivers language without effort—has been compromised. So I have to search manually. I have to consciously hunt through my own memory for something that used to arrive unbidden. The word exists. The system that holds it is intact. What has frayed is the connection between knowing and finding.
The same thing happens when I speak. My brain constructs the sentence correctly. But transmitting some words through my mouth can require more energy than it once did. Under that load, my brain takes the path of least resistance. It drops the middle syllable of a three-syllable word. It slurs across the hard consonants. It is not laziness—it is load-shedding. The system under strain quietly abandons precision to conserve the energy required to keep functioning at all.
What I am describing is not a broken brain. It is a brain whose internal transmission infrastructure has been partially disrupted. My organs are intact. But some connections are frayed.
The System of Systems
Engineers and systems analysts use a specific term for what I am experiencing: a system-of-systems failure. The human body is not a single mechanism but an intricate web of interdependent subsystems—neurological, muscular, cardiovascular, cognitive, linguistic—each capable of functioning independently, yet dependent on reliable pathways of communication with the others.
The critical insight is this: you can have fully functional components and still have a failing system. The failure lives in the connections, not the parts.
This is not a metaphor unique to neurology. It is a structural principle that appears across every complex system we have built—including the ones we depend on for democratic governance and national security.
The Intelligence We Already Had
On the morning of September 11, 2001, the United States government possessed the information it needed to disrupt the attacks. Not all of it, and not assembled into a clear operational picture—but the pieces were there, distributed across agencies that did not share them.
An FBI agent in Phoenix had warned in July 2001 of a possible coordinated effort to train terrorists in American flight schools. The memo never reached the relevant units before the attacks. The CIA had tracked two of the eventual hijackers internationally but failed to transfer that responsibility to the FBI once they entered the United States. The agencies were not broken. The connective pathways between them were.
The 9/11 Commission's central conclusion was not that American intelligence had failed to gather sufficient information. It was that the government had a weak system for processing and using what it already had. The culture of “need to know”—each agency guarding its own data—had severed the connections between fully functional intelligence components. The brain knew the word. It could not transmit it.
The Grid That Would Not Connect
In February 2021, Winter Storm Uri descended on Texas, exposing a vulnerability that engineers and regulators had identified years earlier but chosen not to fix. The Texas power grid—managed by ERCOT—had deliberately isolated itself from the two national interconnections that serve the rest of the continental United States. That isolation was politically protected: connecting to neighboring grids would have triggered federal regulatory oversight, which Texas had structured its energy market specifically to avoid.
When the storm hit, the consequences of that isolation became catastrophic. Natural gas wells lost power and stopped producing fuel, which meant power plants lost their fuel supply, which meant the grid lost generation capacity faster than demand could be shed. Electricity and gas had become mutually dependent in ways that created a negative feedback loop under stress. Within hours, forty percent of grid capacity went offline. ERCOT later determined the system had come within four minutes and thirty-seven seconds of a complete statewide blackout that could have taken weeks to restore.
What failed was not Texas’s power generation capacity. What failed was the connective architecture between systems—between gas supply and power generation, and between the Texas grid and the national transmission infrastructure that could have provided emergency backup. An MIT analysis later found that a properly interconnected Texas grid could have prevented roughly eighty percent of the blackouts during the storm.
The components were adequate. The connections were not.
What We Keep Getting Wrong
There is a pattern across these failures that mirrors what I observe in my own recovery. In each case, the components held under stress. The connections did not.
This is not a failure of awareness. It is a failure of priority. Connective frameworks exist for critical infrastructure— the Cybersecurity and Infrastructure Security Agency and the National Association of Regulatory Utility Commissioners have both developed them. But funding, legislation, and reform consistently flow to the components: more agency capacity, more generation, more redundancy within individual systems. The pathways between systems get attention after catastrophic failure, generally not before.
In governance, connective frameworks exist but remain chronically underdeveloped and underfunded compared to the institutions they are meant to link. In each case, the most overlooked failure was not in the organ. It was in the nerve pathway.
Fixing the Connection, Not Just the Component
What is largely absent is an equivalent framework for governance and political institutions. We have after-action reviews and reorganization legislation. Rarely do we map the connective tissue between institutions, the transmission pathways that allow a functioning center to communicate coherently to its periphery.
Policymakers should stop diagnosing institutional failure at the agency level and start auditing the pathways between agencies—the information-sharing protocols, coordination mechanisms, and transmission infrastructure that connect them. Systems engineers and energy professionals should treat Texas as the canonical case study in interconnection failure, not generation failure. And civic reformers working on democratic resilience should ask not which institutions are failing, but where the transmission pathways between them have degraded beyond their design tolerance. In every domain, the intervention point is the same: the connection, not the component.
The Recovery
I expect to recover. The brain is a remarkable system—plastic, adaptive, capable of building new pathways around damaged ones. The middle syllables are coming back. The pathways are rebuilding.
But the recovery is not happening at the component level. The systems that were always intact—cognition, memory, intention—are not what needed repair. What needed repair were the transmission pathways between them. And that repair is slow and deliberate, requiring a different kind of attention than fixing a broken part.
In rehabilitation, you fix what is broken and rebuild what has frayed. The same logic applies to American institutions and critical infrastructure. Fix what is broken. And do not overlook the pathways. Tend to the connections.
Edward Saltzberg is the Executive Director of the Security and Sustainability Forum, writes The Stability Brief, and is a visiting scholar at George Washington University.
