Historical toponymy and map change

Summary

Beaver Island’s names accumulated; they did not simply replace one another. Anishinaabemowin, French, English, federal survey terminology, Strangite biblical names, Irish settlement names, and modern federal naming policy are all still visible in the record, at different dates and scales.

The held evidence begins with an early Ottawa-associated map label in 1688 and two earlier maps that depict the group without a Beaver/Castor label, then the French I. du Castor in 1744 and 1755 (with I. au Castors on a separate 1755 d’Anville sheet), then moves through Beaver I., Carver’s unlabeled 1778 continental depiction, and Big Beaver I. on English maps, the circa-1849 GLO plats, the 1853 Farmer map, and the 1854–55 Lake Survey chart. The sequence is a history of what mapmakers recorded, not a claim that a map label created local usage.

The Anishinaabemowin layer is attested separately and earlier than any of the English forms in official use: Baraga’s 1853 dictionary prints Amikogenda, Beaver-Island, in Lake Michigan on p. 25.¹

The Strangite map layer

The 1853 Farmer map and the 1855 Lake Survey chart preserve a dense cluster of names associated with the Strangite occupation: Saint James, Beaver Harbour, Enoch, Pool of Siloam, Mount Pisgah, Gennesareth Lake, and Font Lake.²

Names that changed differently

The present Lake Geneserath preserves a Strang-era biblical naming complex, while Galilee and the Jordan survive differently: the Jordan remains a GNIS feature, the former Galilee settlement and township are historical records, and the lake’s spelling has shifted among Galilee, Gennesaret, Genessarett, and Geneserath. Mount Pisgah Bay survives in GNIS as a variant of Donegal Bay, while the dune itself has no equivalent current GNIS feature in the held data.

The 2022 replacement of Squaw Island by Ojibwa Island is a separate modern federal naming action, not a Strangite survival. The gazetteer records the change as a Board decision of 8 September 2022 and keeps the old form as a variant of the same feature.³ Keeping the two changes apart is essential: one belongs to nineteenth-century settlement and religious geography, the other to a twenty-first-century naming-policy change.

How to use the maps

The maps are evidence at different resolutions. A regional map can establish that a name was printed and roughly where a mapmaker placed it; it does not necessarily establish a modern feature identity. A GNIS variant joins a current feature to a historical name, but it does not by itself explain who used the name or when it entered local speech. This survey therefore keeps map images, source articles, GNIS joins, and interpretive themes distinct.

The individual GLO plat-record articles add a more precise federal-survey layer to that comparison. The original sheets preserve labels such as BIG BEAVER I. and GARDEN ISLAND in the circa-1849 survey context. The T39N R09W and T39N R10W supplements must be read separately: their 1943 Pismire Island acceptance and 1954 Section 11 designation are later cartographic actions, not evidence that the names first originated on those dates.

Each plat record also carries GNIS name-record links for the current features inside its modern PLSS extent. That gives a two-step comparison: scan observation to survey extent, then survey extent to current gazetteer row. The modern GNIS attestation stays separate from evidence that a name was printed or spoken in the nineteenth century.

Structured GNIS name-record fan-out

A separate record is held for each of the 111 rows in the GNIS All Names extract for the archipelago. These records preserve exact spellings, feature ids, row classification, and source-citation metadata, and link back to the current place article when one exists.³ They are an index of federal gazetteer evidence, not 111 claims that every form was locally spoken or that every variant was a true name replacement.

Significance

The accumulation pattern this theme documents matters for how older sources are cited here: a nineteenth-century document that uses “Galilee,” “Mount Pisgah,” or “Squaw Island” is not making an error that needs correcting to a modern form — it is using the name current in its own period, and normalizing it silently would erase the very evidence this theme exists to preserve. The held map sequence also shows that map labels and local usage are two different kinds of evidence that are deliberately kept separate here: a name printed on a 1755 sheet documents a mapmaker’s choice, not proof that anyone on the ground used that word, while a GNIS variant record documents a modern gazetteer’s decision to preserve a historical form, not a claim about when or how widely that form was spoken. Two changes in particular illustrate why the distinction between eras must be kept explicit rather than inferred: Strang’s biblical renaming programme of the 1850s and the federal Squaw-Island-to-Ojibwa-Island renaming of 2022 both replaced an island group’s name, using the same island, but 170 years and two entirely different kinds of authority apart. An article that cites one to explain the other, or treats the GNIS gazetteer as a record of when a name was coined rather than when it was last officially adopted, will misdate the change it describes — precisely the confusion this distinction between eras exists to prevent.

Added 1 August 2026 · revised 2 August 2026. Recent changes