If people know anything about geologic time, they know that names like Pleistocene, Jurassic and Devonian are universal concepts that apply to the whole Earth. Only specialists know the real story.
Global Time Terms
The units of geologic time were invented in Europe, by European geologists studying European rocks. The Cambrian Period was first established in the rocks of Wales and given the old Roman name for the area, Cambria. The Jurassic Period was mapped out in rocks of the Jura Mountains in eastern France. The Devonian Period is a hat-tip to Devonshire, England. And so on.
As the geological community expanded overseas and refined its divisions of time, these geographical referents expanded too. The smallest time divisions, the ages, are a gazetteer of obscure places: the Gorstian, the Serpukhovian, the Calabrian, the Chianghsingian. Even though these names honor very specific places, they serve as consistent, worldwide episodes of ancient history. In scientific terms, they are isochronous units like the observations and time signals that synchronize civilization with the rotating Earth.
A few fading regionalisms survive. The single Carboniferous Period in the official world standard is known in America as two periods, the Mississippian and the Pennsylvanian. The Tertiary Period is still recognized by U.S. geologists while the rest of the world splits it into the Paleogene and Neogene. No doubt the geologists of other countries have their own remnant usages that I've never heard of.
Some of the paleontologists are provincialnot personally! I mean that they study the geologic past of geographic provinces, such as continents, and they still have a need for regional terms. The universal time divisions were originally drawn from the fossil history of marine rocks: the extinctions and originations of shellfish, plankton and so on. That's fine for the majority of rocks, which formed in the sea, but not for students of land-dwelling life. And while the seas are one worldwide set of habitats, the land is deeply divided into continents on which life has evolved in separate detailed histories.
Over the course of the 20th century, North American specialists formalized a set of North American Land Mammal Ages that cover Cenozoic time, the last 66 million years. Instead of the familiar sequence Danian, Selandian, Thanetian, Ypresian, . . . , Calabrian, Ionian, Holocene, it goes Puercan, Torrejonian, Tiffanian, Clarkforkian, . . . , Blancan, Irvingtonian, Rancholabrean. The last three of these are named for fossil collections from Mount Blanco, Texas; Irvington, California near San Francisco, and Rancho La Brea, the famous La Brea Tar Pits in Los Angeles.
In Europe, a similar system goes from the Cernaysian to the Villefranchian. The South American version goes from the Tiupampan to the Lujanian, and in Asia it goes from the Gashatan to the Tabenbulakian. (You can find all of these on the Paleobiology Database by searching for time scales on the different continents.) As long as these specialists only talk to each other, everything is fine.
Problems with Provincial Ages
The idea behind the continent-based mammal ages is sound enough: the assemblage of fossil species should be typical of a particular period of time in the geologic past. As in so many fields of knowledge, though, the closer you look the blurrier things may seem.
In North America, a great many fossil localities fall within the two latest time units, Irvingtonian and Rancholabrean. They were originally distinguished, in a 1951 paper by Don Savage, by the presence or absence of bison fossils: Bison are modern animals. The trouble is, the absence of something is not a very robust guideline. A collection of fossils might lack bison just from chance, or because they didn't live in the area at the time. We still don't have a precise idea of when Bison species originated. Exclusively Irvingtonian species include Armbruster's wolf (Canis armbrusteri) and the Irvingon elk (Tetrameryx irvingtonensis), but what if they weren't common enough in the area to show up? What if the habitat was wrong?
Another thing is that, just as North America today is not the same everywhere, the ancient continent was also divided into regions: the mountainous west, the central and eastern plains, the Canadian territory, the Mexican mountains. If bison are the key animal, maybe they couldn't spread from one region to another until they had evolved subtle new species with the right adaptations. That would make these time units diachronous, wandering across time.
And still another thing: the original fossil beds of Irvington, the "type specimen" of the mammal age unit, have been covered up by Interstate 680, so for now we're stuck with what was dug up decades ago and kept in museums. And very few papers have resulted from the original Irvington fossils.
With all this uncertainty since its beginning in 1951, paleontologist Christopher Bell has called the Irvingtonian story "63 years of obfuscation," the result of "historical inertia, competing egos, apathy, ignorance, and inadequate data." That hints at a whole backstage story featuring the human side of science.
Prospects and Perspective
A lot of work in recent decades has gone into the fossil rodents, because mice and their relatives (voles and lemmings and muskrats) are everywhere and evolve rapidly, which makes them useful index fossils like the various microfossil species (such as forams) in marine rocks. Today there are 143 species of these so-called microtine rodents.
A number of rodent-based subdivisions of the Irvingtonian and Rancholabrean have been devised. (Fortunately they don't have names, just Roman numerals.) They are a work in progress: some have poorly defined boundaries and are still being sorted out by region. They promise a more robust system for navigating the Cenozoic, but trying to perfect them within the fuzzier framework of the North American Land Mammal Ages is an awkward way to make progress.
All of this is standard operating procedure in geology. A century ago, the marine fossil specialists were wrestling with the same problems in a somewhat simpler setting. Today the standard time units are tied securely to specific rocks in specific localities: the Global Boundary Stratotype Section and Point (GSSP) or "golden spike" system. Since then the tumult has died down. What we see today in provincial studies is a glimpse at how it used to be everywhere.