"Cognitive Fossils" and the Paleo Mindscape

« previous post | next post »

Below is a guest post by Mark Dow, consisting of an interview with Cory Stade.

Cory Stade is a cognitive archaeologist interested in "how Palaeolithic material culture can inform our understanding of the origin and evolution of language." Formerly a Visiting Fellow at the University of Southampton at the Centre for the Archaeology of Human Origins (CAHO), she received a BA in linguistics from Simon Fraser University in Canada, with a minor in archaeology; a Masters degree in Palaeoanthropology and Palaeolithic Archaeology at University College London; and a PhD in Archaeology at Southampton in 2017 with a thesis addressing "the different modes of culturally transmitted lithic material made experimentally by novice modern knappers." She currently runs an academic proofreading business and, with her partner, a jewelry business called the Stone Age Ceramic Studio.

In her 2020 article "Theory of mind as a proxy for Paleolithic language ability," Stade argues as follows:

The development of "theory of mind" (the term for our ability to infer the mental states of others) in typical human children can be mapped onto the evolution of cognition in humans by examining the stone tool fossil record. This, in turn, can teach us about the evolution of language because of the extent to which theory of mind and language ability are intertwined and predictive of each other.

This interview was conducted via email in the Spring of 2021.

MD: What is your earliest memory or awareness of language? And what first drew you to linguistics?

CS: My earliest awareness of language came not from my own use of it, but through books and reading. My mom had an old tattered ASL guide—she used to work with children with special needs, and many of them used ASL, so she learned it in her early 20s. She taught me finger spelling and some basic words when I was still learning English. My dad studied German in school, and had a textbook, as well as four small leatherbound pocket dictionaries of German, French, Spanish and Italian. I was fascinated by them, and used to carry them around with me. They were very old fashioned, and there was a list of "Christian names" in the back, showing cognate names between the two languages, which I thought was amazing—someone could be Peter in English, but Pierre in French! My dad's name, Ralph, it claimed in German was Rudolph, and I found this wonderful. It opened up the concept that people make different sounds or signs to mean the same thing, and it was captivating.

With my limited resources (I grew up in rural small towns on Vancouver Island, Canada), I was soon drinking up other languages by taking language guides out of the library and watching short shows on learning Japanese on Canada's "The Knowledge Network" station, carefully recording squiggly hiragana symbols in my notebook. I had wanted to attend the French immersion school, but my parents said it was too far away. I still took French classes at school, as well as Spanish and Japanese. My interest was very much in "languages" because I hadn't really learned about what "language" itself was yet.

My first taste of linguistics was when I was 17, and in a Vancouver bookstore. I saw Stephen Pinker's "The Language Instinct," and was hooked right away. Although the theoretical ideas are far from the ones I later developed, it was such an accessible gateway to being introduced to the concept of language and how it operates, how we learn it, and how we learn about it. It's also how I was first exposed to the idea of language evolving in the first place—one of the final chapters is called "The Big Bang" and was very influential in getting my mind whirring with these mysterious questions about how language developed in our species. But it was from that point on that I wanted to study linguistics at university.

MD: What is "language" itself?

CS: Defining language is actually a really difficult task! You want a definition to include everything it is (such as sign languages), but not include anything it isn't (such as animal communication or non-linguistic gestures). There is a chapter at the beginning of Jean Aitchison's book, "The Articulate Mammal," that tackles this amazingly. But to me, language is a learned system of symbols used to intentionally communicate. It took me years to settle on this definition that I'm now pretty happy with.

MD: Does this mean you rule out the possibility of non-human animal languages?

CS: I don't believe there's any evidence to support language existing naturally in any other species today (although there are language trained non-human animals). However, language's origins might very well have originated in our Australopithecine ancestors that you wouldn't really call human. 

Mindscape & Emotional Niche

MD: You write: "Once an individual can attribute others with like mental behaviors, a 'mindscape' is created where thoughts, feelings and knowledge can be shared and understood by others. . . . The attainment of theory of mind therefore permits the very existence of the human social and emotional niche." That's an evocative choice of words. Can you say more about mindscape and niche?

CS: We humans live so much in our minds and the minds of others—we base our decisions off what we think others will think about us, and what we predict they will do. This isn't just limited to one-on-one relationships, but exists in organizations and especially government at every level. This mindscape allows for certain emotions to even exist in the first place, like guilt and sympathy, or shame and jealousy. You can't be embarrassed if you're not able to suppose that people can judge you! And if you're aware that people are thinking about you, as you are of them, suddenly, inside your mind, it's like you're in a room with others. You're no longer alone—you're aware you're being monitored. It's a bit anxiety inducing (it's also the source of our ability to be mentally unwell, especially if we are overattributing thoughts to others, or attributing distorted ones), but it's also a beautiful curse—with theory of mind we're not alone anymore; we can be emotionally supported, empathized with, and understood, and a lot of human personal fulfillment comes from that. I imagine this development would have been even more impactful than the onset of the digital age and people having lives on social media and the internet, which has also just ramped all these social interactions up another notch. And now we have memes, click-bait, fake news, big data, anxiety about impending global disaster; it's all getting very complex in terms of thinking about other's thoughts about others' thoughts; none of this today would be possible without first becoming aware that others "think" in the first place.

MD: You've run experiments in which people, "novice knappers," try toolmaking in various scenarios: by examining a finished product; by watching a video of a toolmaker/teacher work; by in-person, group, silent copying; and through in-person, group, vocal learning. What have these experiments shown?

CS: I was searching for a "signature" that we could look for in Palaeolithic material that showed that complex copying and learning had taken place, so that I could link it to the presence of cognition and language use in the past. I had a hypothesis that I thought was pretty straightforward—those that have similar information about how to make an object are going to end up making similar things. Say, if people have the ability to copy each other's actions (called imitation), they're able to think about the bodily processes that tie the action to the result, or about another's goal. Theory of mind sort of allows for the motivation to copy someone's actions and goal in the first place.

I thought that a community of better copiers are going to be spreading their knowledge around more. And more specific knowledge being spread around about a task is going to result in the objects made from that knowledge being more similar. In terms of stone tools, a community that learns by copying each other are going to have more alike stone tools than a community that learns how to make sharp rocks for tasks in their own way each time. That would provide more opportunity for people to find wildly different ways to do the same thing. And I found in my groups that the more their conditions allowed for the soaking up of the same knowledge about how to make the tool, the more alike the participants' tool shapes came out. And conversely, the groups that learned how to do the task on their own came up with some pretty ingenious (and not always useful) methods, that were pretty diverse!  

So in the Palaeolithic, when there's a number of ways to make a sharp rock which would fulfill the task of butchering an animal, or making a tool, or cutting plant material, and communities are making very specific, finely crafted tools for hundreds of thousands of years—they aren't just coincidentally coming to the same conclusions about how to make it, because that would be way too much of a coincidence. They are making these tools a certain way because that's how they learned to do it.

MD: Can you give an example?

CS: My favorite example is at the site of Boxgrove on the south coast of England, from half a million years ago. Dozens of really fine, thin flint handaxes were found, and a number of them have this really sharp edge made by striking a thin sharp finishing flake off along the top, called a tranchet flake. It's a real "style" that the community of handaxe makers (called Homo heidelbergensis, the ancestor to the Neanderthals) knew how to do well, and it's not simple! I've been knapping for years and I still struggle to thin a handaxe!

These ideas about style led me to believe that a level of standardization in tools, then, in the archaeological record, is a sign of this ability to copy well. This skill ties together the movements of others and their goal, their motivations, that comes with theory of mind. And if you have evidence for theory of mind in the archaeological record, that's where I argue that language skills, which develop alongside and are important in feeding back both ways in their development, can be interpreted, and even measured. It's the idea that a = b = c, a bit of a complicated line of interpretation. But the nature of language means that any interpretation of it in the past is necessarily going to be pretty indirect! And if this hypothesis holds true, then where we see a "style" of stone tool in the archaeological record (and we see "styles" going back hundreds of thousands of years, even millions depending on how complex we see them as), this could be taken as evidence of a certain level of theory of mind and language.

MD: At EVOLANG XI in New Orleans, where we met five years ago, you said: "Language does fossilize—as stone tools in the form of 'cognitive fossils.'"  Are you translating the old mind/body problem into a cognition/stone-tool problem?

Ray Jackendoff described "linguistic fossils" as imprints left in the structure of modern languages of how language used to be and how it emerged. My idea of calling stone tools "cognitive fossils" is the idea that the act of using language, or other cognitive acts, makes impressions on the material world, like a leaf makes its impression in the mud that later becomes shale. There are certain Palaeolithic material traces that just wouldn't exist without certain cognitive abilities having existed in the makers. Complex stone tools are the prime example, because of the copying and even the teaching that would have been required for the knapper to have learned how to produce the tool. 

"Ripples of Energy"

MD: You write that you're considering the when, not the how or why, of language evolution. But what do you imagine as the earliest possible scenario of language? Is there an imagined historical moment in which you hear something familiar?

CS: I think if we start out with the "when," then the "how" and "why" can follow because there will be really different contexts for the lives of people at the time. If language evolved 100,000 years ago, then we're dealing with humans that don't really differ much from human groups today, and we could treat it as a sort of ethnography. But if language evolved 400,000 years ago, or 600,000 years ago, it's a different scenario (a different species!) and many more assumptions we'd take for granted about later periods just can't be made. My short answer is I bet language as we define it began over two million years ago, and more complex languages with "a grammar" evolved over half a million years ago, when stone tools started to become really quite fancy and would have had to have been explicitly taught in order to transmit that "secret" information of how to make them.

To be a little more in depth about it, I imagine the earliest possible birth of language was when Australopithecines, ancestors of the genus Homo, were making early core-and-flake stone tools (called Oldowan, after Oldupai Gorge in Tanzania) two and a half million years ago. People were making tools long before this; the oldest stone tools we know of are older than three million years ago, but they were made in a different, less-procedural way (with indirect percussion, hitting the core against an anvil so sharp useable flakes come off). It's feasible that these oldest tools could have been learned through the ways that chimpanzees learn nut cracking (the physics aren't too dissimilar): the materials are there, others in the group are spending time with the materials and using them, and individuals were putting together the knowledge themselves, not copying.

But these Oldowan tools are more complex. For one, you sometimes get lots of flakes being taken off a single core. This may sound easy, but you can "exhaust" a core pretty easily if you're not thinking about the angles you are making with each hit (you need an acute angle), and you run out of options if it becomes, say, roughly spherical. But I became interested in some refitted flakes and cores at a site called Lokalalei in Kenya. "Refitting" is where you put the flakes back together on the core like a jigsaw puzzle. There are signs on flakes where the impact from the hammer took place, and sometimes even the ripples of energy. Each flake is a bit like a snowflake in shape, so they do slot back together; this means you can be sure you're putting it back together correctly. And there are cores there that archaeologists have reconstructed where tens of flakes have come off a single core. There's a knowledge and sophistication about the knapping there that I don't think resulted from one individual's own learning. I see this as cumulative; these Australopithecines were standing on the shoulders of the knowledge others had passed on, and made in a group that could copy each other.

Imitating, linking another's procedure to their goal, means using a theory of mind, and in child development, having a theory of mind means being able to link symbols to sounds, because you understand what others intend and can copy them. I think that these Oldowan ancestors therefore had a lexicon, and used words to stand for things.

MD: By "lexicon" do you mean words without grammar or syntax?

CS: Yes, a lot of researchers hypothesize an earlier stage of language called "protolanguage," which lacks a grammar but has a repertoire of meaningful sound-meaning mappings. Children speak a sort of "protolanguage" in their early speech, before they learn that we can relate words to each other—not just ordering them serially, but that one word can govern another.

MD: What would you like to know about human language that you don't know?

CS: I think I'm most curious about what Homo erectus communication was like. I think language first arose in Australopithecines, and by the time Homo heidelbergensis was around there was grammar. I don't spend a lot of time wondering about modern humans, or even about Neanderthals; in my opinion Neanderthals would have had a rich, complex language with storytelling and jokes and tongue-twisters, etc. (although, I'd love to hear those stories). But where was Homo erectus in this, two million years ago? This is where my own suppositions are fuzziest. Were they more like the Australopithecines, or did they have a proto-grammar?

If I could get into a time machine, I'd just like to go watch a Homo erectus family for a day—how do they interact? Where do they sleep? What do they eat? Do they wear clothes? Do they have names? What are their relationships like? And are they chattering to each other all the while? I really enjoyed a book I read a few years ago by Ruth Goodman, called, "How to Be a Tudor: A Dawn-to-Dusk Guide to Tudor Life." I think I'd like to sit there and write that book for Homo erectines, and record what the dawn to dusk life of this species was, as it would tell us just so many unknowns about the earliest origins our genus's technology and cognition. I know this question was about language, but I don't think they can be separated; we need to study them all together.

Above is Mark Dow's interview with Cory Stade.


  1. David Marjanović said,

    November 25, 2021 @ 3:46 pm

    My dad's name, Ralph, it claimed in German was Rudolph, and I found this wonderful.

    It's also… wrong, at least in that Ralf and Rudolf (almost never spelled with a spurious ph) are treated as separate today. Without looking it up, I don't think Ralf is a shortening of Rudolf – though Rolf, which also exists, probably is.

  2. KIRINPUTRA said,

    November 27, 2021 @ 2:11 am


  3. Wanda said,

    November 27, 2021 @ 12:17 pm

    I'm confused. Yes, of course you can only get increased sophistication through copying. But why would that have to involve language or instruction? Children can achieve quite sophisticated things merely by prolonged, close observation, and my impression from my (admittedly limited) reading in anthropology is that in pre-scholastic cultures, children usually do learn skills more by observation, imitation, and pitching in than by direct instruction.

    In any case, some of the examples cited in this post (the core with dozens of flakes) strike me as things that any individual, language or not, would have to learn how to do over years of practice. Direct instruction on this type of skill can make the learning process shorter, but procedural memories cannot be directly transmitted through words.

  4. Wanda said,

    November 27, 2021 @ 12:22 pm

    Ok, upon re-reading, I see that the argument is copying :: theory of mind :: language. What's the evidence for the latter relationship? Do deaf-mutes who never learn language incapable of ToM? Or is this saying that in general the cognitive capabilities enable each other? In the latter case, we can only say that, like, these tool-users we're capable of using language, not that they actually did?

  5. Philip Taylor said,

    November 27, 2021 @ 1:17 pm

    Very confused by your last sentence, Wanda ("In the latter case, we can only say that, like, these tool-users we're capable of using language, not that they actually did ?") — is the comma after "like" spurious (and should it therefore be after "users"), and is "we're" to be parsed as "we are" or as a mis-spelling of the past tense of "was" ?

  6. Rodger C said,

    November 28, 2021 @ 10:24 am

    @Philip: I think "like" means "something on the order of" and "we're" should indeed be "were."

  7. Perry said,

    November 28, 2021 @ 5:05 pm

    Interesting. Many species are imitative- not just humans- learning by example. Does this require TOM?
    Also, with complex techniques, learning ( and teaching) can be simpler without words- easier done than said .

RSS feed for comments on this post