This page explains in moderate detail how a particular artifact-bearing deposit -- the most controversial one -- at the Sheguiandah Site in Ontario, Canada, was originally identified as glacial till. Forty years later it was re-interpreted as a beach deposit. It is important to see this issue in its broader context, so for a historical overview, click here.

Why they said it was glacial till

The short answer, for those familiar with tills, is that the controversial deposits were unsorted and contained a majority of rock types foreign to the site. Clasts were often faceted, and the softer types bore glacial striae. The diamictons also contained rounded sandy inclusions. Fabric studies revealed orientations consistent with regional ice flow. Alternatives, such as mudflow and beaches, were explicitly ruled out.
This short answer, however, was long in coming. Geologists visiting Lee's excavations often gave their opinions on sight, but hastily retracted them on learning that artifacts were being found, because the prevailing view that humans entered the Americas after the Ice Age was so strong . Some came back again and again over the four years Lee was excavating, arguing back and forth, raising objections and seeking additional evidence to move forward.

Till is an unsorted sediment, with every size of particle from clay to boulders

We all learn about glacial features in school, such as moraines and drumlins. The material they are made of is called "till". If, as at Sheguiandah, you don't have the landscape features to identify them, you have to dig inside and look for the characteristics of till. At left, a vertical profile of the controversial deposits, overlying another deposit (also identified as till -- see below). Rocks that had been dug up were piled beside the trench.

As glaciers move across a landscape, they pick up material from underneath and carry it along, slowly grinding it ever finer. Enormous boulders are broken down into smaller boulders, cobbles and pebbles. Pebbles are ground into sand grains, and even sand grains are crushed and fractured, producing silt and clay particles. As long as there is a continuous supply of rock underneath, then all stages in the breakdown process, from big to small, will be present in deposits left behind by a melting glacier. Because no particular size class predominates, glacial till is typically referred to as being "unsorted".

Normally, people live in safe and stable places. This seemed to be true at Sheguiandah as far back as the 10,000-year Paleo-Indian level. But as his crews dug below that in what he named the "Habitation Area" (left), Thomas Lee was stumped to see that there were stone tools buried among, and even under, rocks of all sizes (right). He could not believe that people had lived on the spot while these boulders were being dropped all around them. At the end of the first field season, in 1952, visiting geologist John Sanford offered a solution to the puzzle. He suggested that Ice Age glaciers had created the whole deposit, having incidentally picked up pre-existing artifacts. It was, he said, glacial till.
From a geological standpoint, this seemed chronologically possible, because the deposits in question lay below the level at which Paleo-Indian spearpoints had been found. The Paleo-Indians had been the first humans to reach the site after the Ice Age, and they would have been walking on top of fairly fresh glacial till. But it was a problem archaeologically. If the artifacts were already in the ground when the Paleo-Indians arrived, then this was something new and exciting. Could they instead have been made after the Ice Age and then become mixed into the glacial till?

Artifacts --part of the till, or only mixed into it?

Mixing might have happened even while the first Paleo-Indians were still living there. Suppose they had made their first tools on another, higher part of the hill, and left them there. If that piece of ground had then slid or tumbled downhill and come to rest in the Habitation Area, then the situation might appear as it does. The first artifacts, originally on a surface more or less distant, would have been mixed into the underlying deposit as it moved to its final resting place. Then the Paleo-Indians could have resumed work on the new surface, this time leaving behind some of their distinctive spearpoints.
There were several kinds of downhill movement Lee had to consider. Rainwash was ruled out, because it would have left the rocks behind. The deposits were not colluvium, because the nearest slopes were either too distant or too shallow for rocks and soil to have tumbled down by force of gravity. Mud, of course, can flow down lesser slopes and carry even large boulders along, but Lee observed that the sediments he was dealing with wouldn't hold water. As geologist Ernst Antevs summarized the situation, "You can't have mudflow where you can't make mud." In any case, the deposits were were in the middle of a wide, flat area, cut off from higher ground by transverse ridges of bedrock.
Could the artifacts perhaps have been mixed into the glacial till at a later time, long after the Paleo-Indians left? After all, trees do get uprooted, animals burrow, and the expansion of frost heaves the ground.
In one approach to this question, Lee reconsidered the spearpoints, which represent a narrow period of time in archaeological chronology. They were essentially confined to a thin layer about five or six inches down in the soil. The broken halves of some specimens were found at roughly the same depth and in the same stratigraphic level, even when separated by as much as 15 feet (4.5 m). If there had been any substantial mixing, one would expect these artifacts to have become scattered through the soil. It is difficult to argue that non-projectile-points had been mixed downward, while leaving the spearpoints behind. Furthermore, at least two culturally different levels were distinguished below the Paleo-Indian level, and one above it -- a stratigraphic sequence.

The more specific features of till -- rock origins, shapes, and markings -- and lumps of sand

If, then, the artifacts really belong where they were found, how can we be certain that these unsorted sediments are truly glacial till? And how can we be quite sure that the sediments are still where the glacier left them, undisturbed? Specific evidence was sought to answer these questions.
Glaciers, as we have noted, pick up large rocks as they move along. It will often happen that these are carried beyond their native region, and when the glacier drops them, they will be different from the local bedrock in their new location.
The Sheguiandah Site hill is predominately white quartzite, sticking up out of Manitoulin Island's limestone. About half of the stones in the controversial deposits are instead granite and greenstone from more than 20 miles away, across the North Channel -- completely foreign to the site. An additional 22% were shale, which crops out close by, but downhill from the Habitation Area. A continental glacier, irresistibly moving over and around obstacles, was the one force able to move boulders of shale uphill. In that sense, these, too, are foreign to the specific location. Most of the remainder, about 20%, were quartzite and probably local.
In the course of being ground up, several things happen to stones in a glacier. Periodic breakage produces angular fragments, but then continued abrasion rounds off the corners. And within the ice, stones are carried along in first one position, and then flip to another. Abrasion by gritty particles held in the moving ice therefore tends to produce a set of flat, smooth sides -- facets -- with rounded edges. This is evident in the large granite cobbles at right (in a blow-up from the image at the top of this page).
Also, when two rocks held within the ice are ground against each another, the softer one may be scratched. As the stone changes position in the ice, it gets scored in different directions. In Sheguiandah's controversial deposits, both facets and scratches (glacial straie) were specifically noted on the softer shales and greenstones as they came out of the controversial deposits.

In the course of their digging, Lee and his crews were occasionally surprised to find their trowels slicing easily through what they supposed were chunks of disintegrated sandstone. The patch of clean yellow sand at left, beside a small boulder of white quartzite, is one of them (also visible in this page's first picture, right in the center). Another visiting geologist put these finds in a different light. 'What you have there,' he explained, 'are formerly-frozen lumps of sand that a moving glacier ripped out of iron-hard, icy ground and carried along for a time like any other rock.'

Similar lumps of sand occur in a now-petrified till (tillite) in Kluane National Park, Yukon Territory. The one shown at right was evidently melting, with other till materials being forced into it, when its glacier stopped and allowed it to be preserved. Melting can also allow a lump of sand to be smeared out flat before being obliterated. Not only had Lee found ball-like lumps of clean sand in the rather stony till at Sheguiandah, but also sand lenses, about an inch thick and 18 inches across (2.5 x 45 cm). "This," Ernst Antevs advised him, "is your strongest evidence."

Glacial flows and fabrics

But Antevs was by then an old man, and something new and important was coming into the field he had contributed to all his life -- fabrics. No matter how slowly it grinds along, a flowing glacier streams faster in the middle that at the sides, and the lower levels also gradually slide up over those in front. These and other stresses induce an alignment of the stones inside the ice. When the glacier stops and melts away, these patterns of stones may be preserved in the till, rather like the lines of iron filings that remain on a piece of paper after the magnet has been removed. If the deposit is not disturbed, the orientation and strength of the fabric can be measured, even many thousands of years later. A very strong fabric, shown above left, is evident in the same Kluane tillite.

The stresses in a glacier are often revealed by fissures at the brittle surface. In some places, lengthwise stresses predominate, as in the lines of flow in this Kluane glacier (left). In other places on the same glacier, crosswise cracks are more pronounced (right). Within the ice, and in compressed, plastic sediments underneath it, elongated stones tend to line up either with the flow (the more usual case), or transversely, across it.
Both types of alignment were detected in the controversial deposits at Sheguiandah, in different levels. Today, as in the 1950s, the data for a fabric is collected one stone at a time, with a ruler (to determine the length-to-width ratios of the stone), a compass (for bearing in relation to north), and a clinometer (to measure the dip of the longest axis). Computer programs assess the distribution of oriented stones and generate fabric diagrams. At Sheguiandah, Thomas Lee measured nearly 600 stones and plotted the results. He showed that the stones still lay in alignments consistent with the regional flow of glacial ice, which was out of the northeast, toward the southwest (leftmost diagram). Even the transverse orientation, at right, shows an up-ice (northeastward) dip.

Problems with the glacial-till hypothesis

How could tools made by man possibly survive being caught up in a continental glacier? This was the first problem Thomas Lee faced. He eventually came to realize that a hill can protect what lies on the lee, or down-ice side, and keep it from being carried away. It has since been discovered that while glacial ice does flow, it cannot wrap tightly around an obstacle, and leaves an ice-roofed space underneath. At right, Kluane's Kaskawulsh Glacier, flowing around a hill and not quite meeting below it. A lake, dotted with icebergs, fills the gap.

The second problem is that the re-investigation team says there can be only one till and that they have identified it underneath the controversial deposits. Lee's own fabric diagram for this deposit (left) confirms their identification.

The two sets of material give the impression of being quite different. The controversial deposits are looser, sandier, and quite reddish, while the underlying sediment is indurated, looks like clay, and is yellowish-grey. But there are also similarities.

Lithologically, the stones in the two deposits are comparable (47% foreign, 30% shale, and 23% local quartzite in the deeper till, versus 52% foreign, 22% shale, and 20% local quartzite in the controversial tills). Both deposits contain the same kind of microfossils from an adjacent deposit (the Mystic Ridge). And the fabrics for both are similar in strength and direction.

An irregularity in the grain-size curve for the subglacial till led the geologist with the re-investigation to propose that it was "immature," meaning that it was closely related to local material the glacier had just picked up. This argument can be even better made for the controversial deposits, which contain lumps of sand apparently ripped out of a frozen sandy sediment by glacial ice. There need not be two tills, representing two glacial episodes, but two phases of one till, each based on different sediments the ice was overriding at different times..

After the Ice Age, Lake Algonquin's declining shoreline must inevitably have fallen across the entire Sheguiandah Site. Every point on the hill must have been touched by waves at some time. The question is, did the lake make any impression here? Did an ancient beach develop out of glacial till? This is the interpretation of the re-investigation team.