Deep Dive

For decades, Classical Studies professor Nicolle Hirschfeld, Ph.D., has lived an exciting life as an archaeologist. 

She has scuba dived down to ancient shipwrecks in the Mediterranean in the summer, then spent her academic years working to catalog and study the artifacts uncovered by these ventures. 

But as thrilling as these deep dives are, Hirschfeld is equally as excited for the direction this type of research is taking at Trinity. Over the past decade, Hirschfeld has worked to incorporate computer science into her research—first with cataloging the artifacts she uncovered, and now with digitally mapping the sites of wrecked ships.

Hirschfeld at work amidst her findings.

“When ships crash, just think of a container truck that gets dropped onto the highway - it just ‘splats’ on the highway, and everything scatters,” Hirschfeld says. “Because it's a shipwreck, we can find most things and ‘put Humpty Dumpty back together again.’”

And now, with these catalogs and digital maps, all the artifacts Hirschfeld and other archaeologists don’t find at the initial “crash site”—having spilled and drifted further away—are now back in play. And the more complete picture these researchers can get, the better the insights into the economic networks of ancient life become.

“The mapping project is in fact reconstructing how the ship was packed,” Hirschfeld says. “If you can put the ship back together, you know everything that's in the container, but you also can retroactively configure what was packed into each box, and the order in which the boxes were loaded on board, which in turn helps us figure out the route the ship actually took.” 

Hirschfeld (right) and Trinity CS graduate Evan Garvie '16

Whether used for mapping, cataloging or other applications, it’s clear that computer science has become a powerful tool for Hirschfeld and other archeologists. She has regularly collaborated with Trinity computer science professor Seth Fogarty, Ph.D., who specializes in creating domain-specific languages (programming languages dedicated towards a specific area of study) and formal verification (how to automatically prove that a program satisfies a specification).

Over the years, Hirschfeld has engaged with a talented set of Trinity student researchers in the field of computer science.

Hirschfeld and Garvie examine pottery sherds, or fragments

Evan Garvie ’16, now a software engineer at USAA, developed a software program as part of a Mellon Summer Undergraduate Research Fellowship that helped graphically and digitally represent the distribution of pottery sherds found on Hirschfeld’s dive down to a 13th century B.C.E.  vessel. Patrick Bray ’22 is currently updating and improving Hirschfeld’s mapping project.

And Zach Taylor ’20 helped create a revolutionary new research tool for efficiently performing large-scale studies of ancient coins.

Taylor came to Trinity with a love of coin collecting that eventually blossomed into an active research role in the field of numismatics, the study of currency. 

And while Taylor was a computer science major, Trinity’s liberal arts approach allowed him the flexibility to take an introduction to archaeology course with Hirschfeld. Here, he saw the potential to combine his CS skills and love of coin collecting. After Taylor expressed an interest in research, Hirschfeld was able to connect him with the American Numismatic Society (ANS) to start work on a research project that would become his honors thesis.

“At Trinity, computer science students can do either an honors thesis or a senior capstone project,” Taylor says. “I wanted to do research, and I was interested in crafting an interdisciplinary research project. So that's when I approached Dr. Hirschfeld, who hooked me up with Dr. Peter van Alfen at the ANS, and he just so happened to be working on this computer-aided die study (CADS).”

A coin die, Taylor explains, is one of two metallic pieces used to strike (make) a coin. Researchers have been studying them for decades because of the insights that coinage can tell us about ancient society—they are particularly revealing of the methods and scale of production within a particular currency. Know enough about that, and you can work backwards to gain insights about the economic activities of entire city states— in the same spirit as Hirschfeld’s work reconstructing ancient trade routes from the packing methods used in ships. All in all, this is tremendously valuable information for researchers of multiple disciplines, even beyond archaeology, Taylor explains.

Before Taylor’s project, researchers had been using an extremely time-consuming and exhausting process of examining coins by human eye, coin by coin, looking for similarities and differences in each coin to pinpoint which coins came from the same dies. 

But Taylor and Van Alfen, using computer science concepts such as computer vision and hierarchical clustering, developed a prototype process that was able to examine and categorize the similarities and differences between various coins much faster and more accurately than the human eye.

In fact, the subject of Taylor’s honors thesis actually replicated the results of a well known numismatics study that had taken years to complete in a much shorter time frame: “Unsupervised, without any special tuning, our program was able to recreate the original researcher's work overnight,” Taylor says. 

Taylor, left, alongside fellow Stumberg winner Tara Lujan '22

Taylor, who graduated from Trinity University in 2020, is using his CS talents for entrepreneurial pursuits. He is a co-founder of HydroTek (formerly Sapphire), the startup that won Trinity’s 2020 Stumberg New Venture Competition. But his passion for numismatics is still going strong, and thanks in part to his honors thesis, he still holds a title as assistant researcher with the ANS, and is traveling to an international numismatics conference in Warsaw, Poland as part of an invited panel of similar experts.

“If you told 12-year-old Zach that I would be playing with the big dogs for coin collecting and coin research? I’d say that's the coolest thing ever. I love that,” Taylor says. “I think that's something that I can be very proud of, and that was all facilitated by Dr. Hirschfeld.”

Hirschfeld says Trinity is the perfect place for creating the types of interdisciplinary collisions that students like Taylor experience.

“The wonderful thing about Trinity is that we are a great size for a lot of interaction,” Hirschfeld says. “We are small enough that we regularly interact with colleagues from other disciplines, and we’re large enough that we don't exhaust these serendipitous collisions of interest. The faculty is creative and inventive and open, students are encouraged to explore, and the administration finds ways to encourage collaborations through the curriculum [and through] funding support.”

Hirschfeld adds, “Good liberal arts colleges are doing this. I think we're doing it on a level with the best of them.”

And as more CS collaborations unfold, Hirschfeld says these interdisciplinary collisions have the potential to free up student and faculty researchers alike to start thinking bigger—and diving deeper.

“Computer science is super-powering what [archaeologists] used to do, especially methods that we used to do by hand or individually,” Hirschfeld says of archaeological research. “You then have space and time and energy to think more about the bigger picture. Instead of getting buried in the weeds, you see the trees popping up, and then the patterns in which they grow -- freeing up head space and time to think beyond the process of documenting data, to analyzing its patterns.”