Where does interruption collide?

A break that pulls a person away from a task in progress and forces them to set the task aside and later resume it, at a cost. Altmann and Trafton's 2002 Memory for goals modelled why resumption is hard: the goal of the suspended task decays in memory while attention is elsewhere, so returning to it takes time and invites error. Mark, Gudith, and Klocke's 2008 The cost of interrupted work measured that cost in real offices. Here interruption is a lapse and recovery of an internal goal in one mind, not two people talking over each other, a break in electric supply, or a paused course of medicine.[1][2]

Unit

One suspend-and-resume episode on a task, per person

Recovery means

The resumption lag: the extra time and error when the primed goal must be reactivated after the break

Measured by

Timed in the lab by resumption lag and error after a secondary task, and in the field by observation and self-report of switches

In the study of ordinary conversation, an incoming speaker starting up before the current speaker has reached a point where a change of speaker would be orderly, so the start cuts across another's turn rather than following it. Sacks, Schegloff, and Jefferson's 1974 A simplest systematics for the organization of turn-taking for conversation set out the rules of one-at-a-time turn exchange against which such simultaneous talk counts as violative, distinct from a well-timed overlap. Here interruption is a social act between two speakers sharing the floor, measured in the timing of talk, not a goal fading in one head, a loss of power, or a halt in medication.[3]

Unit

One instance of a speaker starting inside another's turn

Recovery means

Not applicable

Measured by

Coded from recorded talk by fine transcription of overlap onset relative to the ongoing turn's structure

In electricity distribution, a loss of service to one or more customers, the event whose frequency and duration define how reliable a network is. The IEEE Guide for Electric Power Distribution Reliability Indices (IEEE Std 1366, 2012 revision) defines the interruption and builds the standard reliability indices, such as average interruption frequency and duration per customer, on top of it. Here interruption is a measurable gap in the delivery of power across a grid, counted per customer and per year, not a lapse of attention, a breach of speaking turns, or a paused drug regimen.[4]

Unit

One loss of supply to a customer, by count and duration

Recovery means

Restoration time: how long until supply returns, aggregated into duration indices such as customer minutes lost

Measured by

Logged by utilities and summarised as indices like SAIFI and SAIDI (interruptions and minutes per customer per year)

In drug therapy, a deliberate stopping and later restarting of a continuous regimen, studied most sharply for antiretroviral treatment of HIV. The SMART trial, reported by El-Sadr and colleagues in 2006 as CD4+ count-guided interruption of antiretroviral treatment, tested pausing therapy when immune counts were high and found the interruption raised the risk of disease and death, ending the strategy. Here interruption is a clinical decision to halt and resume a medication, judged by health outcomes, not a break in a mental task, a violation of conversational turns, or a cut in electric supply.[5]

Unit

One episode of stopping and restarting a drug regimen, per patient

Recovery means

Resumption of the regimen when a threshold is crossed, with outcomes tracked from the pause onward

Measured by

Assessed in trials by clinical endpoints (disease progression, death) and laboratory markers during and after the pause