How we think intersects with the language we think in. Consider the verb in classical Greek, a linguistic tool so complex that it surely allows shadings of thought that are the stuff of finely tuned philosophy. But are the thoughts in our texts genuinely capable of translation? Every now and then I get a glimpse of something integral that just can’t come across in another tongue.

Back in college (and this was a long time ago), I struggled with Greek from the age of Herodotus and then, in the following semester, moved into Homer, whose language was from maybe 300 years earlier. The Odyssey, our text for that semester, is loaded with repetitive phrases – called Homeric epithets – that are memory anchors for the performance of these epics, which were delivered before large crowds by rhapsōdoi (“song-stitchers”). I was never all that great in Homeric Greek, but I do remember getting so familiar with these ‘anchors’ that I was able now and then to read a sequence of five or six lines without a dictionary. But that was a rare event and I never got much better.

The experience convinced me that translation must always be no more than an approximation. A good translation conveys the thought, but the ineffable qualities of individual languages impose their own patina on the words. ‘Wine-dark sea’ is a lovely phrase in English, but when Homer spins it out in Greek, the phrase conjures different feelings within me, and I realize that the more we learn a language, the more we begin to think like its speakers.

My question then as now is how far can we take this? And moving into SETI realms, how much could we learn if we were actually to encounter alien speakers? Is there a possibility of so capturing their language that we could actually begin to think like them?

Let’s talk about Ted Chiang’s wonderful “Story of Your Life,” which was made (and somewhat changed) into the movie Arrival. Here linguist Louise Banks describes to her daughter her work on aliens called heptapods, seven-limbed creatures who are newly arrived on Earth, motives unknown, although they are communicating. Louise goes to work on Heptapod A and Heptapod B, the spoken and written language of the aliens respectively.

Image: A still from Denis Villeneuve’s 2016 film Arrival captures the mystery of deciphering an alien language.

Heptapod B is graphical, and it begins to become apparent that its symbols (semagrams), are put together into montages that represent complete thoughts or events. The aliens appear to experience time in a non-linear way. How can humans relate to that? Strikingly, immersion in this language has powerful effects on those learning it, as Louise explains in the story:

Before I learned how to think in Heptapod B, my memories grew like a column of cigarette ash, laid down by the infinitesimal sliver of combustion that was my consciousness, marking the sequential present. After I learned Heptapod B, new memories fell into place like gigantic blocks , each one measuring years in duration, and though they didn’t arrive in order or land continuously, they soon composed a period of five decades. It is the period during which I knew Heptapod B well enough to think in it, starting during my interviews with Flapper and Raspberry and ending with my death.

Flapper and Raspberry are the human team’s names for the two heptapods they’re dealing with, and we learn that Louise now has ‘memories’ that extend forward as well as back. Or as she goes on to explain:

Usually, Heptapod B affects just my memory; my consciousness crawls along as it did before, a glowing sliver crawling forward in time, the difference being that the ash of memory lies ahead as well as behind: there is no real combustion. But occasionally I have glimpses when Heptapod B truly reigns, and I experience past and future all at once; my consciousness becomes a half-century long ember burning outside time. I perceive – during those glimpses – that entire epoch as a simultaneity. It’s a period encompassing the rest of my life, and the entirety of yours.

The ‘yours’ refers to Louise’s daughter, and the heartbreak of the story is the vision forward. What would you do if you could indeed glimpse the future and see everything that awaited you, even the death of your only child? How would you behave where your consciousness is now, with that child merely a hoped for future being? How would such knowledge, soaked in the surety of the very language you thought in, affect the things you are going to do tomorrow?

A new paper out of Publications of the National Academy of Sciences has been the trigger for these reflections on Chiang’s tale, which I consider among the finest short stories in science fiction history. The paper, with Christian Bentz (Saarland University) as lead author, looks at 40,000 year old artifacts, all of them bearing sequences of geometric signs that had been engraved by early hunter-gatherers in the Aurignacian culture, the first Homo sapiens in central Europe. It was a time of migrations and shifting populations that would have included encounters with the existing Neanderthals.

These hunter-gatherers have left many traces, among which are these fragments that include several thousand geometric signs. What struck me was that these ancient artifacts demonstrate the same complexity as proto-cuneiform script from roughly 3000 BC. Working with Ewa Dutkiewicz (Museum of Prehistory and Early History of the National Museums, Berlin), Bentz notes objects like the ‘Adorant,’ an ivory plaque showing a creature that is half man, half lion. Found in the “Geißenklösterle,” a cave in the Achtal Valley in southern Germany, it’s marked by notches and rows of dots, in much the same way as a carved mammoth tusk from a cave in the Swabian Alb. The researchers see these markings as an early alternative to writing. Says Bentz:

“Our analyses allow us to demonstrate that the sequences of symbols have nothing in common with our modern writing system, which represents spoken languages ​​and has a high information density. On the archaeological finds, however, we have symbols that repeat very frequently – cross, cross, cross, line, line, line – spoken languages ​​do not exhibit these repetitive structures. But our results also show that the hunter-gatherers of the Paleolithic era developed a symbol system with a statistically comparable information density to the earliest proto-cuneiform tablets from ancient Mesopotamia – a full 40,000 years later. The sequences of symbols in proto-cuneiform are equally repetitive; the individual symbols are repeated with comparable frequency. The sequences are comparable in their complexity.”

Image: The so-called “Adorant” from the Geißenklösterle Cave is approximately 40,000 years old. It is a small ivory plaque with an anthropomorphic figure and several rows of notches and dots. The arrangement of these markings suggests a notational system, particularly the rows of dots on the back of the plaque. Credit: © Landesmuseum Württemberg / Hendrik Zwietasch, CC BY 4.0.

As the researchers comment, the result is surprising because you would think early cuneiform would be much closer in structure to modern systems of notation, but here we have, over a period of almost 40,000 years, evidence that such writing changed little since the Paleolithic. Says Bentz: “After that, around 5,000 years ago, a new system emerged relatively suddenly, representing spoken language—and there, of course, we find completely different statistical properties,”

The paper digs into the team’s computer analysis of the Paleolithic symbols, weighing the expression of information there against cuneiform and modern writing as well. It’s clear from the results that humans have been able to encode information into signs and symbols for many millennia, with writing as we know it being one growth from many earlier forms of encoding and sign systems.

We have no extraterrestrials to interrogate, but even with our own species, we have to ask what the experience of people who lived in the Stone Age was like. What were they trying to convey with their complex sequences of symbols? The authors assume they were as cognitively capable as modern humans and I see no reason to doubt that, but how we extract their thought from such symbols remains a mystery to be resolved by future work in archaeology and linguistics.

And I wonder whether Ted Chiang’s story doesn’t tell us something about the experience of going beyond translation into total immersion in an unknown text. How does that change us? Acquiring a new language, even a modern one, subtly changes thought, and I’m also reminded of my mother’s Alzheimer’s, which somehow left her able to acquire Spanish phrases even as she lost the ability to speak in English. I always read to her, and when I tried to teach her some basic Spanish, the experiment was startlingly successful. That attempt left me wondering what parts of the human brain may be affected by full immersion in the language of any future extraterrestrial who may become known to us.

Ludwig Wittgenstein argued that words only map a deeper reality, saying “The limits of my language mean the limits of my world.” Beyond this map, how do we proceed? Perhaps one day SETI will succeed and we will explore that terrain.

The paper is Bentz et al., “Humans 40,000 y ago developed a system of conventional signs,” Publications of the National Academy of Sciences 123 (9) e2520385123. 23 February 2026. Full text. For more on this work, see Bentz and Dutkiewicz’ YouTube video: https://www.youtube.com/@StoneAgeSigns. Thanks to my ever reliable friend Antonio Tavani for sending me information about this paper.