Core Primitive
When a chain breaks restart from the first link rather than trying to jump into the middle.
The phone call at link four
Marco had been running his morning chain for six weeks without a single misfire. Seven links, always the same order: alarm at 6:00, feet on the floor, walk to the bathroom, splash cold water on his face, return to the bedroom, sit at the desk, open the first translation project of the day. The chain was tight. Each link flowed into the next with the kind of automaticity that made the whole sequence feel less like a series of decisions and more like a single continuous motion — which, after six weeks of daily repetition, is exactly what it had become.
On a Wednesday morning, his phone rang while he was standing at the bathroom sink, cold water still dripping from his chin. A client in a different time zone needed an urgent revision. Marco took the call, paced the hallway, resolved the issue in twelve minutes, and hung up. He was standing in the hallway now, not the bathroom. The cold water on his face had dried. The chain had been severed between link four and link five.
His instinct was to pick up where he left off. He walked to the bedroom, sat at the desk, and tried to open the translation project. But the project file sat there on screen, and he did not start working. He checked email. He checked a news site. He opened social media. Forty minutes evaporated before he realized the morning was gone. Marco did not lack discipline that Wednesday. He lacked momentum. The first four links of his chain were not just preparation for the desk — they were the engine that propelled him into the desk. Without the kinetic energy of the preceding sequence, sitting at the desk was an isolated behavior that required the same willpower the chain had been designed to eliminate.
The next morning, after the phone alarm, he started at link one. Feet on the floor. Walk to the bathroom. Cold water. Walk back. Sit. Open the project. Eight minutes, beginning to end. The chain did what chains do when you let the first link fire: it carried him forward.
Why the middle of a chain is not an entry point
The instinct to resume a broken chain from the point of interruption is nearly universal, and it is nearly always wrong. The logic seems sound: you already completed links one through four, so why repeat them? Starting at link five saves time and acknowledges the work already done. But this logic treats the links as independent tasks on a checklist, each one a standalone accomplishment that can be ticked off in any order. Behavioral chains do not work like checklists. They work like dominos.
Each link in a well-constructed chain is a conditioned stimulus for the next link. The alarm triggers feet on the floor. Feet on the floor triggers the walk. The walk triggers the water. The water triggers the return. Each behavior fires because the preceding behavior created the specific sensory and contextual conditions under which the next behavior was learned. In applied behavior analysis, intermediate links function as both conditioned reinforcers and discriminative stimuli — they signal progress toward the terminal reward and set the occasion for the next response in the sequence (Cooper, Heron, & Heward, 2020). Remove the preceding link, and the discriminative stimulus for the next behavior is absent. The behavior can still be performed, but only through deliberate, conscious effort — the very thing the chain was engineered to bypass.
This is where the concept of behavioral momentum becomes essential. John Nevin, whose research on resistance to change spanned four decades, demonstrated that behavioral momentum operates analogously to physical momentum in Newtonian mechanics (Nevin & Grace, 2000). A behavior in motion — a sequence of responses being emitted at a steady rate in the presence of reinforcement — resists disruption. But once the behavior stops, the momentum is gone. It does not pause. It does not wait for you to return. It dissipates. Restarting a sequence after disruption requires rebuilding the momentum from zero, and the most reliable way to rebuild momentum is to begin at the point where the momentum was originally generated: the first link.
Nevin's work showed that the rate of responding after a disruption depends on the reinforcement conditions associated with the context, not on where the disruption occurred. In practical terms, this means that when your morning chain breaks at link four, the relevant question is not "what was link five?" but "what conditions generated the momentum in the first place?" Those conditions live at the beginning of the chain — in the anchor behavior, the physical location, the sensory cues that were present when the first link fired. Returning to those conditions restarts the momentum-generation process.
The warm-up your brain requires
There is a parallel finding in memory research that illuminates why mid-chain entry fails. Researchers studying motor skill performance have long documented what they call "warm-up decrement" — the observation that performance at the start of a practice session is temporarily worse than performance at the end of the previous session, even when no forgetting has occurred (Adams, 1961). The skill is still in memory, but accessing it requires a brief period of reactivation. The first few repetitions of a motor sequence are slightly slower and more effortful than the repetitions that follow. The warm-up period is the brain re-entering the motor program, loading the relevant neural circuits, and re-establishing the timing patterns that make the sequence fluid.
Behavioral chains exhibit the same warm-up requirement. The first link is not just a trigger — it is a warm-up. It activates the neural context in which the entire chain was encoded. Mark Bouton's research on context-dependent learning provides the theoretical framework. Bouton demonstrated that learned behaviors are profoundly sensitive to the context in which they were acquired (Bouton, 2004). Change the context, and the behavior weakens or disappears — not because it was forgotten, but because the retrieval cues are absent. Reinstate the context, and the behavior returns. When you restart a chain from link one, you are not wasting time repeating steps you already completed. You are reinstating the context — the physical location, the body position, the sensory environment — in which the chain was originally learned.
This explains why resuming mid-chain feels labored even when it technically works. The chain runs on willpower rather than momentum. Each subsequent link requires a conscious decision because the contextual reinstatement never occurred. By contrast, when you restart from link one, there is a brief period of conscious execution — maybe the first two or three links feel slightly deliberate — and then automaticity kicks in. The warm-up decrement resolves itself within the first few links, and by the time you reach the point where the chain originally broke, you are moving with the same fluidity you would have had if the interruption never happened.
The restart protocol
Knowing why you should restart is necessary but not sufficient. You also need to know how. A clean restart requires a specific sequence of actions, and defining that sequence in advance — before the chain breaks — removes the decision-making burden at the exact moment when you are least equipped to handle it.
The first step is to accept the break without self-judgment. This is not motivational advice. It is a direct application of Never miss twice's "never miss twice" principle to the chain context. A chain break is a single miss. The break happened. The relevant question is not "why did the chain break" (that analysis comes later, during chain documentation in Chain documentation) but "what do I do in the next thirty seconds?" Self-recrimination does not rebuild momentum. It consumes the cognitive resources you need for the restart. Acknowledge the break the way a pilot acknowledges turbulence — as an event that requires a procedural response, not an emotional one.
The second step is to return to the physical location and body position of link one. If your chain starts at the bathroom sink, go to the bathroom sink. If it starts at your desk, sit at the desk. If it starts with your running shoes by the door, stand at the door with your shoes. You are not just "deciding to restart." You are physically recreating the conditions under which the chain fires. Context reinstatement, as Bouton's research establishes, is not metaphorical. It requires the actual sensory environment that was present during encoding. The more precisely you recreate the starting conditions, the more reliably the chain will re-ignite.
The third step is to execute the first link. Not to think about executing it. Not to plan to execute it in five minutes. Execute it now. The first link is your anchor — the behavior you designed, per Chain anchors, to be the strongest and most reliable element in the chain. It fires without willpower. It fires regardless of mood. Let it fire.
The fourth step is to let momentum carry you forward. Once the first link completes, the second link's discriminative stimulus is present. Let the second link fire. Then the third. Do not think about the fact that you already did these steps earlier. Do not evaluate whether they are "necessary." They are necessary — not as tasks to be completed, but as momentum-generators that prime each subsequent link. The chain is a sequential motor program, and it needs to run from the beginning to achieve its characteristic fluidity.
The fifth step addresses a practical reality: sometimes the chain breaks and there is not enough time to run the full sequence. In this situation, the answer is not to skip links — skipping middle links severs the discriminative stimulus chain just as thoroughly as the original interruption did. The answer is to run a shortened version that preserves the sequential structure: link one, link two, and the terminal link. You fire the anchor, build enough momentum to engage the automation, and jump to the ending that closes the loop and delivers the terminal reward. This is the core concept behind emergency chains, which Emergency chains will develop fully. For now, the principle is simple: a short chain that runs from beginning to end is always superior to a full-length chain that starts in the middle.
Breaks are data, not verdicts
There is a crucial reframe embedded in the restart protocol that deserves explicit attention: a chain break is information about your chain's design, not evidence of your personal inadequacy. When Marco's chain broke because of a phone call, the break revealed a design vulnerability — the chain had no recovery protocol for external interruptions during the early links. That is a structural observation, not a character judgment. The chain was under-engineered for the conditions of Marco's actual life, where clients call at unpredictable hours. The fix is structural: silence the phone during the chain, or design the chain to tolerate brief interruptions at specific points, or pre-build the restart protocol so that recovery is itself automated.
This reframe matters because the emotional response to a chain break determines whether you restart or abandon. If the break triggers shame, the shame activates avoidance, and avoidance means the chain does not fire the next day or the day after that. Never miss twice's abstinence violation effect is in play: one break becomes permission to stop. But if the break triggers curiosity — "the chain is vulnerable at link four when external interruptions occur; how do I fix that?" — then the break becomes a diagnostic event that improves long-term resilience. The restart protocol operationalizes the curiosity framing: you execute the restart procedure now and analyze the break later, when the chain is safely running again.
Over time, if you track your chain breaks (which Chain documentation will help you do through documentation), you will notice patterns. Breaks cluster at specific links, at specific times of day, under specific conditions. These patterns are engineering specifications for a more robust chain. A chain that has broken and been rebuilt multiple times is stronger than one that has never broken, because its weak points have been identified and reinforced.
The Third Brain
An AI assistant is particularly well-suited to support chain rebuilding because it can hold two perspectives simultaneously: the procedural perspective (what to do right now) and the analytical perspective (what the break pattern means over time). In the moment of a chain break, you need the procedural perspective. You need someone — or something — to say: "Go back to link one. Start the chain. Analyze later." An AI can deliver that instruction without the emotional coloring that your own internal voice adds. Your internal voice says "you broke the chain again." The AI says "restart protocol: return to the bathroom sink."
Beyond the immediate restart, you can use an AI to maintain a chain-break log. Each time a chain breaks, report three data points: which link broke, what caused the break, and what time of day it occurred. After ten or fifteen break events, ask the AI to analyze the log for patterns. You may discover that 80 percent of your breaks happen at the same link, which means that link needs structural reinforcement per Chain strength depends on the weakest link. You may discover that breaks cluster on specific days of the week, suggesting an environmental variable you had not considered. You may discover that external interruptions have a higher recovery rate than internal resistance, which tells you something important about the difference between a design problem and a motivation problem.
You can also ask an AI to help design shortened emergency chains for time-constrained restarts. Describe your full chain, specify how much time you have, and ask it to identify which links are structurally essential and which can be compressed while preserving sequential integrity. The AI can propose emergency variants at different time thresholds — a five-minute version, a ten-minute version, a two-minute version — so that you have pre-built recovery options ready before you need them.
The chain you can see
You now have a restart protocol for when chains break: accept the break without judgment, return to the starting position, execute link one, let momentum rebuild, and use a shortened chain if time is limited. You understand why restarting from the beginning works — behavioral momentum must be rebuilt from the point of generation, context reinstatement requires the original sensory conditions, and warm-up decrement resolves only through sequential execution. And you understand that breaks are diagnostic data, not character verdicts.
But effective restarting depends on knowing exactly what your chain looks like — every link, every transition, every potential vulnerability. Most people carry their chains in implicit memory, running them automatically without ever articulating the sequence explicitly. This works fine when the chain is running smoothly. It becomes a serious problem when you need to rebuild, because you cannot restart a sequence you cannot describe. The next lesson addresses this directly: chain documentation — the practice of writing out your chains in explicit, link-by-link detail so that the structure is visible, auditable, and repairable.
Sources:
- Cooper, J. O., Heron, T. E., & Heward, W. L. (2020). Applied Behavior Analysis (3rd ed.). Pearson.
- Nevin, J. A., & Grace, R. C. (2000). "Behavioral Momentum and the Law of Effect." Behavioral and Brain Sciences, 23(1), 73-130.
- Adams, J. A. (1961). "The Second Facet of Forgetting: A Review of Warm-Up Decrement." Journal of Experimental Psychology, 62(3), 257-273.
- Bouton, M. E. (2004). "Context and Behavioral Processes in Extinction." Learning & Memory, 11(5), 485-494.
Frequently Asked Questions