Segmentation-Dependent Event Counts: Why How Many Depends on Where One Ends

This conceptual paper introduces Segmentation-Dependent Event Counts, an operation in the Representation and Measurement series of On Human Modeling. Before anyone can count events, a rule must establish where one ends and the next begins. Two observers can count the same activity accurately and reach different totals because they draw different boundaries. The paper asks whether the unit fits the purpose of the count before asking whether anyone tallied it correctly. The two volumes of Kill Bill reached theaters in 2003 and 2004. Quentin Tarantino counted them as one film because he had written and made them as one work. The distinction mattered to his announced plan to retire after ten features. When Once Upon a Time in Hollywood appeared in 2019, it was his ninth feature by his own count and his tenth if each theatrical volume counted separately. After the destruction of the World Trade Center on September 11, 2001, insurance coverage turned on what counted as one occurrence. The insurers together had provided about $3.5 billion in coverage per occurrence. Under the broker’s form, which grouped losses attributable to one cause or a series of similar causes, the destruction counted as one occurrence. A jury applied two-occurrence treatment to nine other insurers. The same events yielded different counts under the terms that bound each insurer, with consequences for how much the insured parties could recover. The paper places the operation alongside Jeffrey Zacks and Barbara Tversky’s account of how people perceive continuous activity at finer and coarser scales. The choice of scale also matters when an AI assistant counts interruptions in a meeting recording. Suppose a listener speaks over a colleague four times while the colleague tries to finish one point. A team studying overlapping speech would count four overlaps. A team studying how often speakers are kept from finishing a point could count one sustained interruption in the same recording.

Authors

Publication Details

Journal
Zenodo (CERN European Organization for Nuclear Research)
Published
2026-11-21
DOI
https://doi.org/10.5281/zenodo.24409181
Primary Topic
Categorization, perception, and language
Type
article
Field-Weighted Citation Impact
0.00
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
article

Segmentation-Dependent Event Counts: Why How Many Depends on Where One Ends

Alexandra Cary
Zenodo (CERN European Organization for Nuclear Research)
Categorization, perception, and language
article

Segmentation-Dependent Event Counts: Why How Many Depends on Where One Ends

Alexandra Cary
article en

Abstract

This conceptual paper introduces Segmentation-Dependent Event Counts, an operation in the Representation and Measurement series of On Human Modeling. Before anyone can count events, a rule must establish where one ends and the next begins. Two observers can count the same activity accurately and reach different totals because they draw different boundaries. The paper asks whether the unit fits the purpose of the count before asking whether anyone tallied it correctly. The two volumes of Kill Bill reached theaters in 2003 and 2004. Quentin Tarantino counted them as one film because he had written and made them as one work. The distinction mattered to his announced plan to retire after ten features. When Once Upon a Time in Hollywood appeared in 2019, it was his ninth feature by his own count and his tenth if each theatrical volume counted separately. After the destruction of the World Trade Center on September 11, 2001, insurance coverage turned on what counted as one occurrence. The insurers together had provided about $3.5 billion in coverage per occurrence. Under the broker’s form, which grouped losses attributable to one cause or a series of similar causes, the destruction counted as one occurrence. A jury applied two-occurrence treatment to nine other insurers. The same events yielded different counts under the terms that bound each insurer, with consequences for how much the insured parties could recover. The paper places the operation alongside Jeffrey Zacks and Barbara Tversky’s account of how people perceive continuous activity at finer and coarser scales. The choice of scale also matters when an AI assistant counts interruptions in a meeting recording. Suppose a listener speaks over a colleague four times while the colleague tries to finish one point. A team studying overlapping speech would count four overlaps. A team studying how often speakers are kept from finishing a point could count one sustained interruption in the same recording.

Zenodo (CERN European Organization for Nuclear Research)
Categorization, perception, and language
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon