Human reading looks so ordinary that it is easy to treat it as the default way brains extract meaning from marks. Yet when I look at what animals can do with symbols, numbers, and sounds, the picture that emerges is far stranger than a simple “they cannot read” verdict. Animals do not process written language as humans do, but they routinely crack symbolic codes, juggle abstract concepts, and even exploit our writing in ways that force us to rethink what reading really is.
The more closely I follow the research, the more it feels as if humans built one very specific highway through the landscape of communication, while other species have carved out their own intricate side roads. Those paths do not lead to novels or street signs, but they do reveal flexible minds that can map symbols to the world, sometimes with a precision that looks uncannily like our own literacy.
What scientists actually mean by “reading”
When people ask whether animals can read, they usually have a very human scene in mind: a person silently scanning a page, turning squiggles into words, and words into ideas. In cognitive science, that skill is defined as a layered process that links visual patterns to sounds, then to vocabulary, and finally to complex meaning. By that strict standard, current evidence shows that animals do not “read” the way humans do, even when they learn to respond to printed symbols or screens full of icons, a point that researchers underline when they describe how bonobos and dolphins decode complex symbols without ever acquiring full human-style literacy, as recent work on bonobos and dolphins makes clear.
That distinction matters because it separates the mechanics of recognizing marks from the deeper architecture of language. Human reading is built on a foundation of spoken language that is already rich in grammar, metaphor, and narrative, so the written code is essentially a second layer of abstraction. When scientists say animals do not read, they are not denying that other species can match shapes to outcomes or that they can learn elaborate symbol systems. They are saying that, so far, no nonhuman animal has been shown to use visual marks to access an open-ended, creative language in the way a literate person does when moving through a novel, a legal contract, or a social media feed.
Symbol-savvy animals and the limits of comparison
Even if animals fall short of human literacy, their ability to handle symbols is far from trivial. In controlled experiments, primates, birds, and even some marine mammals have learned to associate arbitrary shapes with objects, actions, or rewards, then to combine those shapes in ways that suggest a grasp of rules rather than simple rote memorization. I find it striking that bonobos and dolphins can navigate symbol boards with dozens of options, selecting icons to request food, toys, or social contact, which shows that they can treat visual tokens as stand-ins for real-world categories rather than as mere pictures, a capacity that recent work on complex symbols highlights.
At the same time, the comparison to human reading can be misleading if it ignores how much training and structure go into these animal systems. Where a child eventually picks up new written words in the wild, without explicit rewards, most animal symbol studies rely on painstaking conditioning and tightly controlled contexts. The animals excel within those boundaries, but they do not spontaneously generalize their symbol knowledge to new domains in the way literate humans do when they encounter unfamiliar signage, brands, or digital interfaces. That gap suggests that while many species can master sophisticated symbol use, the leap from symbol to fully flexible language remains a uniquely human step.
Monkeys that treat numerals like abstract tools
Some of the clearest evidence that animals can handle abstract symbols comes from number research. In a series of experiments, scientists have shown that monkeys can learn to associate Arabic numerals with specific quantities of items, then use those numerals to make choices that reflect an understanding of “more” and “less.” I am struck by how the presence of numerals seems to sharpen their performance: when monkeys are given the option to choose between sets labeled with symbols, their accuracy in picking the larger quantity improves, which suggests that the symbols themselves help them think more clearly about quantity, a pattern documented in work where researchers note that, although chimps, monkeys, and parrots can link Arabic numerals to numbers of items.
What fascinates me is that this looks less like memorizing a trick and more like using a cognitive tool. Once the monkeys have internalized that “5” stands for a certain magnitude, they can compare “5” and “7” without physically counting dots, which mirrors how humans offload mental work onto written symbols. The fact that their choice accuracy jumps when numerals are present hints that the symbol system is not just a passive label but an active aid to reasoning. That is not reading in the literary sense, but it is a form of symbolic thinking that narrows the gap between human and nonhuman minds more than many people expect.
The symbolic monkey and the price of tokens
Numbers are not the only domain where animals reveal a knack for abstraction. In another line of research, tufted capuchin monkeys have been trained to use tokens as stand-ins for food, effectively turning pieces of metal or plastic into a kind of currency. I find it revealing that these capuchins can learn that one token “buys” a certain amount of food while another token buys more, then adjust their choices to maximize their payoff, which shows that they are not just reacting to immediate sensory cues but are tracking the symbolic value of the tokens, a result that underpins the description of capuchins in work titled The Symbolic Monkey.
What makes these findings so provocative is that the monkeys sometimes fall for the same irrational biases that humans show in economic experiments, such as overvaluing certain token types or misjudging trade-offs when the options are framed differently. That parallel suggests that once a brain starts treating objects as symbols for value, it inherits both the power and the pitfalls of abstraction. The capuchins are not reading price tags in a supermarket aisle, but they are navigating a miniature economy where meaning is detached from physical form, which is one of the core moves that made human writing and money possible.
Pigeons, pattern-spotting, and the illusion of spelling
Not all animal symbol work happens in high-tech labs or with charismatic primates. Pigeons, often dismissed as urban background noise, have turned out to be surprisingly adept at visual categorization. In one widely discussed example, trained pigeons can distinguish between correctly spelled and misspelled strings of letters, pecking at the “real” words and ignoring the fakes. When I look at those results, I see a powerful reminder that pattern recognition alone can mimic aspects of reading, since the birds are not accessing meaning but are still sensitive to the statistical structure of letter combinations, a point that has filtered into public discussion where people note that Pigeons can “read” and spot misspellings.
From my perspective, the pigeon data highlight both the power and the limits of equating symbol discrimination with literacy. The birds are exquisitely tuned to visual regularities, which is a crucial part of what human readers do when they learn which letter sequences are likely in a given language. Yet the pigeons do not attach those patterns to concepts like “dog” or “food,” nor do they use them to build sentences. Their performance shows that a brain can become a sophisticated statistical engine for shapes without ever crossing the threshold into language. That distinction matters when we interpret claims that animals “understand” written language, because it reminds us that recognizing a pattern and grasping its meaning are related but separable skills.
Why human language still stands apart
To understand why animal symbol skills do not quite add up to reading, I find it helpful to zoom out to the broader question of language. Human language is not just a set of signals but a system that allows speakers to generate and understand an unlimited number of new sentences, including ones they have never heard before. Researchers who compare human and nonhuman communication emphasize that many animal signals are tied to immediate contexts and fixed responses, while human language is creative and unpredictable, a difference captured in work that notes that the fundamental gap between human and non-human communication is that animals tend to react instinctively, while human language is creative and unpredictable.
That creativity is what written language plugs into. When I read a sentence, I am not just decoding letters, I am drawing on a mental grammar that lets me interpret nested clauses, metaphors, and hypotheticals. Animals that learn symbols can map them to specific outcomes or categories, but they have not been shown to use those symbols to build open-ended structures with the same combinatorial freedom. This is why many linguists argue that, even if we someday teach an animal to recognize hundreds of printed signs, we would still be missing the deeper ingredient that makes reading a gateway to literature, law, and science rather than a clever parlor trick.
Language as a rule-based design system
Another way I think about the human advantage is through the lens of design. Some philosophers and cognitive scientists describe language as the systematic use of designators, which means that it relies on rule-based manipulation of abstract signs rather than on direct emotional or reflexive signals. In that view, what sets human communication apart is not just vocabulary size but the way we can detach signs from immediate contexts and recombine them according to learned rules, a point that is made explicitly in work that defines Language as the systematic use of designators.
Reading, on this account, is a specialized extension of that rule-based system into the visual domain. When a child learns that the letters “c-a-t” stand for a particular sound and concept, they are mastering a design rule that can be applied to countless other words. Animals that learn symbols often do so in narrower ways, with each symbol tied to a specific reward or object rather than to a generative pattern. The fact that some species can stretch beyond that, as in the case of monkeys using numerals or tokens, suggests that the capacity for rule-based symbol use is more widespread than once thought. Yet the full-blown design system that lets humans discuss the weather forecast, legal codes, or fictional universes still appears to be a uniquely human construction.
AI is listening in on the animal world
While the cognitive gap between human reading and animal symbol use remains, technology is starting to change how we listen to other species. In California, an effort based in Berkeley is using artificial intelligence to analyze vast libraries of animal sounds, from whale songs to bird calls, in the hope of decoding their structure and meaning. I find it telling that this project treats animal communication as a complex data problem, feeding recordings into machine learning systems that can pick out patterns too subtle for human ears, a strategy that has been described in coverage of An East Bay company that hopes to decode animal language.
From my perspective, this AI work does not blur the line between human reading and animal communication so much as it reframes the question. Instead of asking whether animals can learn our written codes, researchers are asking whether we can learn theirs, using algorithms as translators. If those systems eventually reveal that some species use more structured, combinatorial signals than we realized, it could force a rethink of how unique human language really is. Even then, though, the fact that we need advanced computation to parse those patterns underscores how specialized our own literacy is: we are the species that not only reads but also builds machines to read the world on our behalf.
More from Morning Overview