Pilot experiment: can we distinguish the stone tool products of flaking from shaping?

I was collecting data from my pilot experiment around this time last year. The experiment tested whether there were measurable differences between the stone tool products of flaking and shaping. Turns out, it's complicated (go figure) - but I did find some significant differences in the flakes! I'm still working on some of the data a year later, but I wanted to share a little bit of the experimental process and results with you today.

The first step in this experimental process was to collect the appropriate rocks. Luckily, I made friends with a smart geologist at Stony Brook who was willing to take me on a road trip. Steven searched a geological database and found a source of phonolite rocks in northern New Jersey that had similar mineral composition to phonolite rocks in northern Kenya. Weird, but cool! We drove for a few hours, came upon a well-maintained hiking path, and stumbled upon the original source of the New Jersey phonolite that we were looking for.

 

We found an outcropping of the original phonolite source.

We found some cobbles - but it looks like so did other folks - they were used as building materials in this historic home!

 

We knew there were phonolite cobbles close by because we found some that were used to build this historic home!

 

This is a typical photo of me in the field - dragging luggage full of rocks.

 

I collected measurements of the cobbles' shapes before they were broken apart in the experiment. Eventually, I plan on comparing the variation in the natural cobble shapes to potentially differences in shape variability in the products of the two experimental conditions (flaking vs. shaping).

 

The experiment was designed around two conditions:

1) Flaking: The toolmaker was instructed to create as many flakes as possible, but they weren't allowed to control the shape of the cobble as they went along. They had to rotate the cobble after each flake they removed. This constraint made sure the toolmaker could not shape the core of plan each subsequent flake removal.

2) Shaping: The toolmaker was instructed to shape the cobble into an long and symmetrical tool. They had to maintain this shape as they continued to remove flakes from the core.

I had some awesome assistants help me collect each flake after it was produced, label it with a number, put it in a sample bag and set it aside. This made it easy for data recording later on because I knew at what stage each flake came from (the beginning, middle, or end of the sequence). The participant completed each condition seven times (that number isn't special, it's just how many repetitions I could complete based on the number of cobbles Steven and I were able to collect in New Jersey).

The next part of the experiment is pretty boring - I measured and analyzed over 500 flakes! I use a computerized data recording system to collect all of the data from the stone tools. Digital calipers with USB cables and computer software help speed up this process. I typically record nearly 20 different data points (both qualitative traits and quantitative measurements) for each flake, so anything that speeds up data collection is a bonus.

 

I compared the data from the flakes in the flaking vs. shaping conditions and two measurements showed significant differences: flake internal platform angle and flake platform thickness. The platform is the part of the flake where the toolmaker initially hit the stone. This part of the flake is sensitive to the specific angle and force that the toolmaker applied to the stone. It's really important to collect measurements from these parts of flakes because they help us quantify differences in toolmaking actions.

 

I found significant differences in flake internal platform angle in flakes that were made during early stages of the reduction sequence.

It's pretty cool when you find some significant differences - that means flaking and shaping toolmaking processes can produce measurably different stone tools, and maybe these measurements could be helpful for understanding variation within archaeological stone tool assemblages.

However, this experiment only tested differences between two very specific types of toolmaking processes, and it can be hard to know what toolmaking processes you are looking for in the archaeological record.

That's why my new project is using a single set of experimental flaking baseline data to compare to the archaeological stone tools. This allows the archaeological data to differ in a wide range of ways (not just assessing whether the archaeological data fits into one category of flaking/shaping or another). This is a more open-ended approach that may identify several different toolmaking behaviors from the past.