+++
title = "The Boundary Lost the Vote"
description = "A named era makes a gradual or uneven transformation look like a single dated break. Chronology and periodization provide the apparatus for testing whether the boundary belongs to the evidence or to the analyst."
date = 2026-08-19
draft = false

[taxonomies]
topics = ["chronology", "historical-eras"]
kinds = ["convention", "attestation"]

[extra]
tier = "public"
schema_type = "Article"
canonical = ""
hero = "articles/the-boundary-lost-the-vote/hero.webp"
hero_alt = "Chronological table split by a red archival boundary strip"
diagram = "diagrams/the-boundary-lost-the-vote.svg"
podcast = ""
mechanism = "Chronology separates an absolute anchor from a relative sequence, while periodization records a label, its boundary rule, and its geographic scope; a period name can organize evidence without identifying a causal discontinuity."
discriminating_test = "Shift the boundary across plausible dated anchors and regions; a result that survives the shifts has evidence beyond the canonical label date, while one that exists only there is classification-sensitive."
false_friend = "A genuine discontinuity also aligns with a date, but it is accompanied by an independently measured change in rules, records, or mechanism rather than by a convenient period name alone."
symptoms = [
  "A regime chart shows a sharp factor break at an analyst-chosen historical date",
  "A long backtest is divided into eras after the returns are already known",
  "A global market phase is named from one region's institutional history",
  "A result disappears when the proposed boundary moves to another plausible date",
]
sources_verified = false

[[extra.faq]]
q = "How can an analyst tell whether a regime break is real?"
a = "An analyst moves the proposed boundary across every plausible anchor and reports the resulting grid. A break that survives those movements has evidence beyond one classification date; a break that appears only at the canonical date is boundary-sensitive. The Assyrian sequence becomes absolute because the Bur-Sagale eclipse matches 15 June 763 BCE, not because a later historian gave its surrounding years a name."

[[extra.faq]]
q = "Why is a historical era a weak way to divide a backtest?"
a = "A historical era can have several defensible boundaries and still remain useful for explanation. The Gilded Age begins variously in 1865, 1873, or 1877 and ends in the 1890s or at the 1896 election; its label came from Twain and Warner's 1873 novel but historians adopted it in the 1920s. A backtest partition chosen from that label can manufacture the contrast it reports."

[[extra.faq]]
q = "Can a global market phase be named from one country's history?"
a = "It can be named that way, but the name does not establish global applicability. The Gilded Age describes the United States, the Victorian era describes Britain and its empire, and the Cold War describes a US-Soviet bipolar order. The three-age system likewise loses utility in Pre-Columbian America and Australia. A phase label needs an explicit geographic scope before it can support a cross-market comparison."

[[extra.faq]]
q = "What makes a dated break more credible than a convenient period name?"
a = "A credible dated break has an independent record that fixes the date and a mechanism that changes with it. Rawlinson's 1867 identification of the near-total eclipse visible over northern Mesopotamia tied an Assyrian eponym year to 15 June 763 BCE. By contrast, the Venus Tablet repeats visibility configurations on cycles of roughly 8, 56, and 64 years, so it leaves several candidate dates rather than one break."

[[extra.faq]]
q = "How should a regime boundary be reported when the date is disputed?"
a = "A regime report states the candidate dates, the evidence for each date, the rule that selects one, and the geographic scope, then shows results for each plausible boundary. The Anthropocene proposal selected Crawford Lake in July 2023 for a 1950s plutonium signal, but the ratifying subcommission rejected formal adoption in March 2024 by 12 votes to 4, with 2 abstentions. A named period can remain useful without becoming a settled date rule."

[[extra.sources]]
title = "\"Assyrian eclipse,\" Wikipedia"
publisher = "Wikimedia Foundation"
url = "https://en.wikipedia.org/wiki/Assyrian_eclipse"
kind = "tertiary"
fetched = "2026-08-13"

[[extra.sources]]
title = "\"Chronology of the ancient Near East,\" Wikipedia"
publisher = "Wikimedia Foundation"
url = "https://en.wikipedia.org/wiki/Chronology_of_the_ancient_Near_East"
kind = "tertiary"
fetched = "2026-08-13"

[[extra.sources]]
title = "\"Egyptian chronology,\" Wikipedia"
publisher = "Wikimedia Foundation"
url = "https://en.wikipedia.org/wiki/Egyptian_chronology"
kind = "tertiary"
fetched = "2026-08-13"

[[extra.sources]]
title = "\"Eponym dating system,\" Wikipedia"
publisher = "Wikimedia Foundation"
url = "https://en.wikipedia.org/wiki/Eponym_dating_system"
kind = "tertiary"
fetched = "2026-08-13"

[[extra.sources]]
title = "\"Hallstatt plateau,\" Wikipedia"
publisher = "Wikimedia Foundation"
url = "https://en.wikipedia.org/wiki/Hallstatt_plateau"
kind = "tertiary"
fetched = "2026-08-13"

[[extra.sources]]
title = "Wikipedia, \"Anthropocene Working Group\""
publisher = "Wikimedia Foundation"
url = "https://en.wikipedia.org/wiki/Anthropocene_Working_Group"
kind = "tertiary"
fetched = "2026-08-13"

[[extra.sources]]
title = "Wikipedia, \"Anthropocene\""
publisher = "Wikimedia Foundation"
url = "https://en.wikipedia.org/wiki/Anthropocene"
kind = "tertiary"
fetched = "2026-08-13"

[[extra.sources]]
title = "Wikipedia, \"Bronze Age\""
publisher = "Wikimedia Foundation"
url = "https://en.wikipedia.org/wiki/Bronze_Age"
kind = "tertiary"
fetched = "2026-08-13"

[[extra.sources]]
title = "Wikipedia, \"Christian Jürgensen Thomsen\""
publisher = "Wikimedia Foundation"
url = "https://en.wikipedia.org/wiki/Christian_Jürgensen_Thomsen"
kind = "tertiary"
fetched = "2026-08-13"

[[extra.sources]]
title = "Wikipedia, \"Cold War\""
publisher = "Wikimedia Foundation"
url = "https://en.wikipedia.org/wiki/Cold_War"
kind = "tertiary"
fetched = "2026-08-13"

[[extra.sources]]
title = "Wikipedia, \"Dark Ages (historiography)\""
publisher = "Wikimedia Foundation"
url = "https://en.wikipedia.org/wiki/Dark_Ages_(historiography"
kind = "tertiary"
fetched = "2026-08-13"

[[extra.sources]]
title = "Wikipedia, \"Gilded Age\""
publisher = "Wikimedia Foundation"
url = "https://en.wikipedia.org/wiki/Gilded_Age"
kind = "tertiary"
fetched = "2026-08-13"

[[extra.sources]]
title = "Wikipedia, \"Iron Age\""
publisher = "Wikimedia Foundation"
url = "https://en.wikipedia.org/wiki/Iron_Age"
kind = "tertiary"
fetched = "2026-08-13"

[[extra.sources]]
title = "Wikipedia, \"Jacob Burckhardt\""
publisher = "Wikimedia Foundation"
url = "https://en.wikipedia.org/wiki/Jacob_Burckhardt"
kind = "tertiary"
fetched = "2026-08-13"

[[extra.sources]]
title = "Wikipedia, \"Jules Michelet\""
publisher = "Wikimedia Foundation"
url = "https://en.wikipedia.org/wiki/Jules_Michelet"
kind = "tertiary"
fetched = "2026-08-13"

[[extra.sources]]
title = "Wikipedia, \"Late antiquity\""
publisher = "Wikimedia Foundation"
url = "https://en.wikipedia.org/wiki/Late_antiquity"
kind = "tertiary"
fetched = "2026-08-13"

[[extra.sources]]
title = "Wikipedia, \"Middle Ages\""
publisher = "Wikimedia Foundation"
url = "https://en.wikipedia.org/wiki/Middle_Ages"
kind = "tertiary"
fetched = "2026-08-13"

[[extra.sources]]
title = "Wikipedia, \"Renaissance\""
publisher = "Wikimedia Foundation"
url = "https://en.wikipedia.org/wiki/Renaissance"
kind = "tertiary"
fetched = "2026-08-13"

[[extra.sources]]
title = "Wikipedia, \"Three-age system\""
publisher = "Wikimedia Foundation"
url = "https://en.wikipedia.org/wiki/Three-age_system"
kind = "tertiary"
fetched = "2026-08-13"

[[extra.sources]]
title = "Wikipedia, \"Victorian era\""
publisher = "Wikimedia Foundation"
url = "https://en.wikipedia.org/wiki/Victorian_era"
kind = "tertiary"
fetched = "2026-08-13"
+++

In 1873 Mark Twain and Charles Dudley Warner published *The Gilded Age: A Tale of Today*. The title eventually named a period of United States history. It never dated one.

The beginning is placed at the end of the Civil War in 1865, at the Panic of 1873, or after Reconstruction in 1877. The end goes broadly into the 1890s, or at the presidential election of 1896.

The phrase itself did not enter historical writing until Van Wyck Brooks, Lewis Mumford and Charles Austin Beard picked it up in the 1920s, roughly fifty years after the novel it came from.

None of those choices is careless. Each makes a different claim about what mattered: a war's end, a panic, the close of Reconstruction, an election, or a retrospective literary description applied long afterwards. The trouble starts when a chart gives the label one date, treats that date as an event, and then discovers a sharp regime change exactly where the label was placed.

## The name arrived fifty years after the period

The Gilded Age is unusually candid about the gap between a coinage and a periodization. Twain and Warner supplied a title in 1873; historians supplied the era in the 1920s. The conventional span runs about twenty years, and it can mean 1877 to 1896 or 1865 to 1896 without either use breaking a rule.

There is no settled rule to break.

Other era names hide the same plurality behind much cleaner edges. The Victorian era is bounded by an accession and a death. The Cold War convention runs from a doctrine speech in March 1947 to Soviet dissolution in December 1991.

Those boundaries rest on dated political records, and the records are real.

They do not establish that any particular institution changed character on the morning the record was made, which is the claim a regime chart actually needs.

The Renaissance makes the larger version of the point. Burckhardt's 1860 account made the rupture from the Middle Ages canonical, and Michelet applied the name to a whole era. Panofsky answered that the two periods were linked by a thousand ties. Nobody in that dispute doubts that fifteenth-century Italy existed. The argument is over whether the word identifies a break or organises a continuity, and the two produce opposite charts from identical evidence.

A label can be an excellent filing system and an unreliable change-point estimator.

Those are not in tension. They are different instruments, and the label is only ever the first one.

## A boundary carries a jurisdiction whether or not it says so

Christian Jürgensen Thomsen set out the three-age terminology in 1836, from artifact assemblages: stone, then bronze, then iron. The sequence has little utility in Pre-Columbian America or Australia, where other frameworks are used instead, and its travel outside northern Europe is where the trouble shows.

Sub-Saharan Africa is the damaging case. Iron smelting is documented at Lejja in Nigeria and at Obui in the Central African Republic well into the second millennium BCE, with Nok smelting by the middle of the first millennium and Djenné-Djenno producing iron a few centuries later.

The sequence never supplies the regional Bronze Age the label presupposes.

An era name imported from one region can describe a real local process and manufacture a false ordering everywhere else it is applied.

The name does not record where it was calibrated.

## The vote that refused a boundary

The Anthropocene is the clearest recent case, because the refusal was formal and the reasoning was written down.

The working group developed its proposal from 2009, selected Crawford Lake in Ontario as the reference site in July 2023, and proposed a 1950s start marked by plutonium from hydrogen-bomb testing. In March 2024 the ratifying subcommission of the International Union of Geological Sciences declined to adopt it, by a wide margin.

The boundary lost the vote.

The stated objection was not that human impact was unreal or too small.

It was that the changes happened at different times in different places, which defeats a single stratigraphic boundary.

That is a body voting against a boundary it agrees describes something.

The term stays useful in science and public discussion; it does not enter the formal time scale, and the two decisions were kept apart deliberately.

A global market phase named from one market has precisely this defect. The evidence may support a genuine local sequence while the imported boundary does nothing but create synchrony in every other market it is applied to.

## Explanation and allocation are different jobs

Historical eras survive fuzzy borders because their work is explanatory. The Gilded Age gathers a disputed run of American history into a name that can be argued about even while its first and last years remain unsettled. The value is in the organisation. There is no promise attached that a measurable process changed on one particular morning.

A trading or risk regime has a different job, and the difference is not a matter of rigour. Once a regime selects exposures, it needs a boundary rule that reproduces and a record of the evidence that chose it.

A declared set of plausible anchors can be tested against each other. A named era chosen after the returns are known cannot be distinguished from the result it produces.

The Gilded Age is a serviceable name for a contested stretch of American history, and it costs nothing when the dates move.

A regime boundary that moves capital has to be an anchor that survives being moved, and the only way to find out is to move it first.