PALEOLITHIC ARTIFICIAL MEMORY SYSTEMS:

We have long considered ancient rock art including cave paintings to be data. I also include carved bone and ivory as well as portable carved and modified stones. The authors of this new paper have a modified way of looking at these records. They refer to them as Artificial Memory Systems (AMS), suggesting that after they had been created they could be visited again later to retrieve data from them.

While they focused on bone and ivory it seems to me that their approach could just as well be used looking at painted caves, or any other rock art. “Artificial Memory Systems (AMS) encompass devices that record, store, transmit, and retrieve coded information beyond the brain, via external representations. AMS can be anything from the notches on a gunslinger's pistol, tracking past success, to the symbols on and data encoded within the Voyager spacecraft's golden record, detailing a snapshot of Earthling knowledge and culture.” (Jackson 2025) This is all true, although it is a pretty broad field of examples.

 A. and B. – marks on bone made by modern butchery. From Courtenay et al., 2025,  Figure 1, page 3.

The methodology the team used will be to analyze two groups of marked artifacts, the Paleolithic and what they refer to as ethnographic. The ethnographic artifacts are much more recent and we have actual ethnographic data about their meanings. “Current scientific knowledge suggests humans are the only species to manufacture and use these tools. While a number of artifacts dating back to the Middle Paleolithic have been considered to be early instances of AMS, conclusive and systematic evidence of this function is absent. Here we contrast the spatial distribution of markings on these potential early AMSs to other Paleolithic artifacts displaying butchery and ornamental marks, as well as ethnographically recorded cases of AMS. We find that both eth­nographic and Upper Paleolithic AMSs are endowed with systematically different signatures that distinguish them from the other artifacts. These findings suggest that modern humans in at least Africa and Europe had sophisticated cognitive capabilities for information storage and retrieval, providing insights into the possible development of quantity-related cognition.” (Courtenay et al. 2025) Many researchers have considered the question of numeration in these examples of ancient art, from dots and lines among the animals on cave walls, to grooves on a boulder as a given but the question of what they do or do not count is constant.

 

C.and D. – decorative engraving in the form of zoomorphic (C) and geometric (D) motifs. From Courtenay et al., 2025,  Figure 1, page 3.

Statistical analysis was applied to the Paleolithic bone markings. “The present analysis has shown that the spatial distribution of different types of markings on bone are separable, with distinct patterns emerging for butchery activities, figurative or abstract representations, and potential AMSs; whether the latter be ethnographic examples, or Paleolithic instances interpreted as such from previous studies. These studies have proposed that four distinct factors, in isolation or com­bination, may play a role in creating codes allowing for the storage of information in an AMS; the number of marks, the accumulation of marks over time, their spatial organisation and arrangement, as well as their morphology. To date, the identification of potential Paleolithic AMSs has been based mainly on the technological analysis of marks.” (Courtenay et al. 2025)

E.  and F. – possible AMS. From Courtenay et al., 2025,  Figure 1, page 3.

Having used statistical methods to analyze the Paleolithic examples, the team then used the same methods on more recent artifacts for which ethnographic records are available. “The examined ethnographic datasets included examples of notched sticks, some of which exhibit quantity-related representations, and some that strongly suggest some form of expressive quantification. They are documented from 20th -century artefacts from Muacapenda and Muatchondo (Angola), tally sticks from Medieval England, as well as 19th -century notched wooden artifacts from the Yakun­bura community of the Dawson river (Australia). This raises a plausible interpretation for these artifacts, related with the emergence of quantitative cognition in our species. Although these artifacts present no additional discernible information about their function based solely on the mark­ings beyond the data that has been collected ethnographi­cally or historically, our statistical analyses show that they share many properties with the potential Paleolithic AMSs in our sample. An important remaining question, however, is to determine what aspects of expressive quantification might be involved in Paleolithic AMSs, since quantifica­tion is a multifaceted phenomenon that encompasses a wide range of features and cognitive mechanisms.” (Courtenay et al. 2025) Comparing the results of their study of both groups of artifacts, the team concluded that the Paleolithic markings were indeed capable of recording data, but since we have no knowledge of their languages it is unlikely that we will ever be able to retrieve that data.

G. and H. – known AMS. From Courtenay et al., 2025,  Figure 1, page 3.

Of course we are still using AMS, we just have very technologically different systems nowadays. “Writing systems, recorded music and images stored digitally in computer code represent the most modern form of AMS, utilizing the most complex mix of language and technology ever created, while essentially relying on the simplest binary system of notches and dots to encode.” (Jackson 2025) Of course writing systems, recorded music and digital images rely on considerably more complicated tools to achieve their function. Indeed, I am typing this into a Lenovo laptop computer. My actions in this are not really more complicated than carving notches in a bone would be, but I am only able to do this because someone went to considerable trouble to manufacture my tool, much more complicated than a rock flake.

So, do these markings encode data? If the obvious fact that the marks were produced on purpose by someone with an idea behind them can be considered data, then yes, they do. I really don’t know what this gets us, however. We already believed this. It sort of strikes me as arguing about the difference between red and scarlet. I suppose it is comforting to know that their statistics back up what we believed all along.

REFERENCES:

Courtenay, Lloyd Austin, Frencesco d’Errico, Rafail Nunez, Damian E. Blasi, 2025, Identifying potential palaeolithic artificial memory systems via Spatial statistics: Implications for the origin of quantification, Archaeological and Anthropological Sciences 17 (171). https://doi.org.10.1007/s12520-025-02286-4. Accessed online 5 August 2025.

Jackson, Justin, 2025, Searching for Artificial Memory Systems in ancient humans with spatial statistics, 5 August 2025,  Phys.org online. https://phys.org/news/2025-08-artificial-memory-ancient-humans-spatial.html. Accessed online 5 August 2025.