Did Prehistoric Puerto Ricans Make Clam Chowder?

Not finding cooking technology from 2,500 years ago, archaeologists tap isotope analysis to investigate how exactly the pre-Colombian people of Puerto Rico prepared their shellfish

I am clam. Hear me squeak.
I am clam. Hear me squeak.Credit: febb
Ruth Schuster
Ruth Schuster
Ruth Schuster
Ruth Schuster

So, did prehistoric Puerto Ricans use clay pots to boil clams? Inquiring minds want to know — and now an international team led by Dr. Philip Staudigel of Cardiff University has enlightened us.

Apparently, they did not, the team reported Wednesday in Science Advances.

That conclusion is based partly on the absence of pottery remnants from the period at the site between the fifth century B.C.E. to the first century C.E., and chiefly on analysis of how the shellfish were cooked. It seems, based on analysis of the shells, the pre-Colombian people of Puerto Rico ate their clams barbecued.

“Our chief purpose was to learn more about cooking methods of the past,” says Staudigal, explaining the goal of the study to Haaretz. “Later inhabitants of Puerto Rico commonly cooked their food in ceramic vessels as soups. … The conclusion based on our measurements, and the absence of ceramics near the site, is that they probably didn’t have ceramic technology yet [before the Common Era]. Our results provide an independent, albeit indirect, validation of this.”

Puerto Rico has been inhabited for at least 6,000 years, but the earliest settlers in the Caribbean seem to have been pre-pottery hunter-gatherers. So while ceramic firing apparently began as early as 20,000 years ago in China, and certainly existed in Japan about 18,000 years ago, the technique of shaping clay into vessels and firing them was slow to reach the Caribbean.

Indications for how the pre-Colombian people of Puerto Rico prepared their Phacoides pectinatus strengthens the hypothesis that ceramic technology was a late arrival to the islands.

It is true that by about 500 B.C.E., pottery had reached nearby islands in the Lesser Antilles, brought there by immigrants from continental South America, probably Venezuela, Staudigel tells Haaretz. (This population would later be replaced by the Taíno culture, who were the Puerto Rican natives when Columbus arrived, he adds.) Some pottery fragments found there predate these migrations: there may have been trade between the islands and mainland, Staudigel says.

In pre-Common Era Puerto Rico specifically, whether or not they managed to get their hands on some ceramic pots we haven’t found, it seems clams were roasted on an open fire. So no, they did not make clam soup.

The charm of clam

The discovery that pre-Colombian Caribbean residents ate clams is not a surprise. Probably all ancient peoples living by the sea ate shellfish and anything else they could trap, net, spear, scavenge, collect or pick up that died on the beach. This has been going on since time immemorial. A Homo erectus may have eaten the clam before engraving its shell in Java half a million years ago — the earliest engraving found so far. Archaeologists excavating caves on South Africa’s Pinnacle Point found remains of edible shellfish dating to about 165,000 years ago and evidence has been found that Neanderthals also appreciated shellfish. They’re easy to catch (bend over, pick up) and they’re nutritious.

But the question was how the early Puerto Ricans cooked them.

Our story starts with a pre-Colombian shell midden (garbage pile) discovered in Cabo Rojo, eastern Puerto Rico, right by the coast. The researchers extracted 24 specimens of the toothsome bivalves from the midden, from layers dating to about 500 B.C.E. to 1 C.E., and tested them to elucidate how they were processed. The control group was raw clams collected on the beach.

This is a clam.
This is a clam. People eat this.Credit: Allen Lew

The boiling point of freshwater is 100 degrees Celsius (212 degrees Fahrenheit), and that of seawater is 103 degrees Celsius. Once boiling, soup also stays at around 100 degrees Celsius, though the specific temperature depends on the composition of the soup. Fire, however, would heat clams to temperatures beyond 100 degrees Celsius and up to 200 degrees Celsius.

Clumped isotopic analysis revealed that half of the Cabo Rojo shells had been heated to temperatures beyond 100 degrees Celsius and up to 200 degrees Celsius. In the other half of the shells, no such indication was detected.

“The clumped isotope technique measures the abundance of rare bonds between specific isotopes of carbon and oxygen in certain minerals,” Staudigel helpfully explains. “As minerals are heated, these bonds can be rearranged. Experiments have shown that as mineral aragonite, which makes up many marine organisms, is heated above 125 degrees Celsius, these bonds can break and re-form, in a manner proportional to the intensity and duration of heating. By measuring the abundance of these bonds in the clams from an archaeological site, workers can infer the peak heating temperature during cooking.”

He stresses that this technique cannot differentiate between raw shells and boiled ones, or barbecued shells that didn’t reach a temperature their technique could detect.

Analyzing a sample clam in the lab.
Analyzing a sample clam in the lab.Credit: Philip Staudigel
Hamhung clam bake North Korean style
Clam bake, North Korean style, which could well be how the pre-Colombian Puerto Ricans cooked their shellfish.Credit: Uri Tours

The bottom line is that half the shells were heated well over 100 degrees Celsius and half weren’t. So: The clamshells heated beyond 100 degrees Celsius had likely been roasted on an open fire. The other half could have been from clams eaten raw; boiled somehow; fire-grilled but only to low temperature; or, plausibly, the ancient Puerto Ricans simply placed their clams onto the fire and didn’t flip them. The bottom shell facing the fire or sitting on the hot surface would have heated between 100 to 200 degrees Celsius, and the top one would have remained relatively cool.

Think about it: If you drop a clam into boiling water, it will cook evenly. Its two half-shells will reach the same temperature. But if you bung a clam on a barbecue and don’t flip it, only the bottom shell on the hot surface will be seriously heated.

“When I barbecue clams, I place them on the grill. When they open, I season the flesh and eat them. This method of cooking could result in the asymmetric heating we are describing,” Staudigel says. Clumped isotopic analysis by a future archaeologist would postulate that either he only roasted half the clams, or that he roasted them all but only from one side.

Wouldn’t a clambake cause telltale char marks on the shells? “Probably, but when shells are in the dirt for 2,000 years they all look a bit dirty,” Staudigel answers.

Clams on fire

Though the possibility sounds far-fetched, the team investigated whether the unfortunate gastropods could have somehow become embroiled, so to speak, in wildfire.

Since clams are not known for their terrestrial mobility, the heated shells would have had to originate in islander meals of raw (or boiled) clams, after which the shells were heated by wildfires to 100-200 degrees Celsius.

Excavating the prehistoric site in Puerto Rico
Excavating the prehistoric site of Cabo Rojo in Puerto RicoCredit: Jessica Joseph

That could have happened. But no direct evidence of wildfires was found at the excavation: no ash horizons, no chalky burned horizons of shells, etc., Staudigel says. Also, the heat from a wildfire would be stratigraphically contained. Burned shells would have been in one layer, not scattered throughout the occupation period of the site. But scattered they were; the bivalve specimens were taken from different layers dating from about 500 B.C.E. to 1 C.E.

Also: The prediction for wildfires heating the deposit of shells and sundry detritus to 100-200 degrees Celsius would be for all shells to be uniformly warmed by the heated earth. This was not observed, he explains. Finally, roasting the bivalves wouldn’t require the use of ceramic technologies (and, indeed, none were found at the site).

And thus, the archaeologists reached a conclusion about how people cooked clams 2,000 to 2,500 years ago without having found their pots.

“The benefit of our method is that it only examines the residual material from the clams and infers the peak heating temperature — this allows for them to be interrogated independently of or in the absence of other archaeological data,” the researcher explains. “This enables investigations into ancient cooking methods in circumstances where there is no preserved cooking technology and all that remains is the discarded table scraps.”

He adds that clumped isotope analysis is a technique still very much in its infancy: This publication was the first deliberate investigation of an archaeological site using clumped isotopes to determine ancient cooking methods.

Asked what else the prehistoric Puerto Ricans ate, Staudigel says the evidence shows that this group ate bivalves and marine gastropods, suggesting they subsisted mostly from the sea. “It is likely they also fished. But if they did, there was poor preservation of fish bones,” he says. “The site was likely occupied periodically over its 800 years of activity, so what the people ate when not at this location may have varied depending on what was available.”

Their cooking methods aren’t the only mystery surrounding the pre-Colombian population of Puerto Rico. Haaretz reported in July on a collection of unique and enigmatic figurines given to a monk by their dying guardian in the late 1870s. These could be artifacts from a lost civilization that lived in Puerto Rico as long as 3,000 years ago, according to Prof. Reniel Rodríguez Ramos of the University of Puerto Rico at Utuado.

One may surmise that this lost civilization didn’t have pottery skills, either. Widespread use of ceramics only reached the island during the Arawak migration, Staudigel says. And thus, the beneficence of soup would reach Puerto Rico too.