Without checking, I have to think that I carefully couched my wording there. For that audience, I did use the more famous example of Stonehenge, for which there is some pretty solid evidence for my claim. There are other examples in the archaeological record that certainly are older and in other cultural contexts. I can easily think of standing stones that are as old as 12,000 years old for which the same case could be made in borderline agricultural societies. The tough part is that would have required the definition of standing stones and a lot of other pieces which I didn’t feel I had the time to create the context for in that setting.
I imagine that there are potentially examples of this sort of behavior going back as far as 30-40,000 years or more, but there is is no direct (known) archaeological evidence left to make such cases. There are oral histories of indigenous peoples in Australia that indicate memories of things that do exist in the geological record to provide some evidence of this.
I’ll also point out that astronomical use is NOT equal to memory use. To make that claim you’d need a lot of additional evidence. In fact, I might suggest something stronger, particularly about Stonehenge. Stonehenge’s primary use was not an astronomical one. Its primary use was as a mnemonic device. The astronomical one was important for the ritual practice (we would call it spaced repetition in modern psychology and pedagogic contexts), but wholly tangential.
If you’re interested in the underlying evidence, Dr. Lynne Kelly has an excellent Ph.D. thesis on the topic, but you might find her book The Memory Code, which expands on the thesis, more accessible. She’s also got a great bibliography of these topics on her website.