Some fossils are “problematic.” They resist all attempts at classification. They lack a home on the tree of life.

The Tully monster, Tullimonstrom gregarium, is one such problematic creature.

“This animal doesn’t fit easy classification because it’s so weird,” says Lauren Sallan, an assistant professor in the Department of Earth & Environmental Science in the School of Arts & Sciences. “It has these eyes that are on stalks, and it has this pincer at the end of a long proboscis, and there’s even disagreement about which way is up.”

That’s why Sallan and many of her colleagues were surprised when, last year, two groups of scientists claimed they had solid evidence that the Tully monster, which lived 300 million years ago, was in fact a vertebrate; specifically, a fish.

Upon closer examination of these claims, Sallan and a group of colleagues found them flawed. Writing in the journal Palaeontology, they provide alternative explanations for the other researchers’ observations.

“Tully is anything but a fish,” Sallan says.

Thousands of Tully fossils have been found in central Illinois. Tully monsters are torpedo shaped and around a foot long. Previously, they were believed to be a worm, a mollusk, or even a lobster-like arthropod.

Yet one of the 2016 studies claimed to have determined that a light band running down the midline of some of the creatures was not a gut tube, as scientists once believed, but instead a primitive backbone known as a notochord. Using this as evidence that the Tully monster was actually a vertebrate, that study’s authors went on to note that other structures, such as gill sacs and teeth, confirmed this classification, and placed it as a close relative to living lampreys, a type of jawless fish.

But Sallan and colleagues noted several incongruities in the study authors’ evidence, among them the fact that the researchers failed to take into account that internal structures, such as a notochord and gill sacs, wouldn’t have been preserved in the marine deposits at Mazon Creek due to high acidity in the area.

The other group suggested Tully monsters were vertebrates based on scanning electron microscope images of the creatures’ eyes. They observed structures called melanosomes, which are typically found in vertebrates.

Yet the Penn-led team argue that this, too, does not hold up as solid evidence that Tully is a vertebrate, noting that the “cup” shape of its eyes more closely resembles those of primitive chordates, molluscs, and worms.

Sallan says that not only do new analyses claim to find things that are unlikely to be preserved in the Mazon Creek’s conditions, they fail to find other structures universal to vertebrates—such as sand-like grains in the inner ear that help animals balance—that she has found in other vertebrate species in the same deposits.

And thus, the Tully monster case is still an open question.

“If you’re going to make extraordinary claims, you need extraordinary evidence,” Sallan says.