Here’s a test. Without looking them up on Google, try to guess the meanings of the foreign antonyms tobi and kekere. They are words in Yoruba, a widely spoken West African language that has its roots in the Old Stone Age. The words are equivalent to the English antonyms big, small. Now, take another guess: which one’s which?

The majority of us will arrive at the same answer. In fact, if we repeat the exercise with antonyms for shapes, sound intensities or even brightness in other foreign languages, we will still agree more than half the time. The trend holds even for non-existent words. In a famous linguistic test, subjects almost always gravitate to the non-word baluma to describe rounded shapes and takete for more angular objects. If you think about it, there appears to be something inherently rounded about baluma,and sharp and pointed about takete. Likewise, in the Yoruban example above, tobi seems like an apt choice to depict bigobjects whereas kekere is more fitting for smaller entities. In other words, the dimensions are ostensibly encoded in the sound of the words.

Human languages are rich in words that sound like what they mean and the sound-meaning associations, or sound symbolism, are surprisingly similar across languages. But how does our brain link phonetics with meaning? In a recent study, psychologists at the University of Rochester, NY and the University of Sussex, United Kingdom, showed that people with synesthesia – a condition where a stimulus, such as sound, can evoke unusual perceptions of color, taste, or odor – are especially adept at matching unknown, foreign words with their meanings. Their work offers clues to the origins of sound symbolism, and explains why we develop intuitions about the meanings of words we have never heard before.

Synesthesia, which occurs in about 1% of the population, arises when an increased number of nerve fibers interconnect discrete regions of the brain, causing more than average “cross-wirings.” Synesthetes, as these individuals are called, lead normal, healthy lives except that they experience additional sensations to sensory stimuli, viz. colors or tastes for words, touch for sounds, and so on. For 56-year old British synesthete James Wannerton, for example, words and names spontaneously evoke distinct taste sensations in his mouth. He coined his own version of the London underground map, in which the stations are named after the tastes they trigger on his tongue – King’s Cross to him tastes like a moist Dundee cake. Wannerton’s name-taste associations, which have lasted all his life, presumably arise from a cross-talk between word processing and taste centers of his brain.

In sound symbolism, the sound of words can bring images to our mind – think of pequeño, petit, or kleine (meaning small) as opposed to grande, grand, or gross (meaning large). Although sound symbolism is more implicit and less involuntary than synesthesia (we need clues to be able to make the right guesses), it can still be thought of as a process that cross-activates multiple – auditory and visual – areas of the brain.

In the present study, psychologists Kaitlyn Bankieris and Julia Simner tested the assumption that sound symbolism stems from similar cross-wirings in the brain as seen in synesthesia. The scientists recruited native English-speaking controls and grapheme-color synesthetes, who see colors in letters or numbers (graphemes) – for example, yellow for the letter c, red for the number 4, and so forth. They asked the subjects to listen to an audio recording and guess the meanings of hundreds of foreign words, given two alternatives. The words spanned four semantic groups, big/small, bright/dark, up/down and loud/quiet, and included vocabulary from ten different languages – for instance, jhiinu (Gujarati, for small), olimikka (Tamil, for bright), neerwaarts (Dutch, for down), among others.

First, the scientists found that both groups of participants were remarkably good at deciphering the meanings of the foreign words albeit only in the big/small and loud/quiet semantic domains. Their performance was barely better than chance in the up/down or bright/dark categories.

To some extent, this discrepancy reflects how our mind maps sounds to meanings. In domains such as big/small, sound symbolic words translate aspects of size to physical aspects of the vocal tract, a linguistic feature termed iconicity. When we say grand (French, for large), for example, our mouth expands as if to mimic the size of the object we refer to, whereas, when we say petit,the vocal tract constricts, and the word plants an impression of a tiny object. Sound symbolic words similarly capture sound intensities but it gets trickier to maintain iconicity when denoting directions or brightness. As one theory has it, words in these domains are encoded by such facets of speech as intonation but because the playback in the study consisted of words spoken in a neutral tone, the subjects faltered in the latter areas.

The scientists next arrived at a more intriguing finding. In the former word categories, where both participants excelled, they discovered that the synesthetes outshone the controls in guessing the meanings of the unknown words.

This was certainly not because they had a stronger vocabulary per se or a superior IQ – the synesthetes were just about as good as the controls in an English word power test. Instead, they seemed to have the edge because of their better grasp of sound-meaning associations in the foreign words. It is tempting to postulate that their increased sensitivity to sound symbolism, a process that links auditory and visual senses, emerges from the cross-wirings seen in synesthesia.

Indeed, psychologists Romke Rouw, Vilayanur Ramachandran and others have unveiled unusual cross-connections in synesthesia. Using a technique called diffusion tensor imaging, scientists have traced the course of nerve fibers in grapheme-color synesthetes and found surprisingly dense connections near the fusiform gyrus, a tubular brain structure that sits just above the ear. The fusiform gyrus contains a site, where graphemes are perceived, abutting the color processing area V4. The consensus is that the dense connections in the synesthetes are fibers transgressing the boundary between adjacent grapheme and color areas. On a behavioral level, these extraneous wirings cause a mix-up of perceptions, inducing colors for letters and numbers.

In synesthetes, scientists in fact observe multiple hubs of such cross-connections in various other parts of the cerebral cortex. An exciting avenue for future research is to find out whether sound symbolism, like synesthesia, taps into one or more of these cross-connections.

Of course, sound-meaning maps are not exclusive to synesthesia. If we recall the experiment, the synesthetes surpassed the controls in the same word categories in which the controls fared better than chance. This implies that if the synesthetes made their choices by pursuing certain word-meaning associations, perhaps by virtue of their cross-wirings, the non-synesthetes made similar associations but based on their gut. They were only less efficient than their counterparts. This leads us to the fascinating speculation that the cross-wirings seen in synesthetes are but an extreme version of a cross-talk that is in fact ingrained in all of us.

In support of this idea, an expanding trove of evidence from child psychologists and linguists suggests that as infants, we all start out with cross-wired brains, much like synesthetes, and are equally alert to sound symbolism in all languages. But as we specialize in our native language, these cross-connections wither away and we grow out of our sensitivity to foreign languages. Scientists postulate that in synesthetes, on the other hand, the cross-wirings persist into adulthood, due to genetic mutations that interfere with the pruning process.

This so-called “childhood synesthesia hypothesis” is compelling because it means that we all retain an indelible imprint of the early cross-wirings, which is why we make the same kind of associations in later life. And just as we are able to guess the meaning of foreign words today, we may have used these shared intuitions to help invent language. Perhaps a nascent vocabulary of sound symbolic words was mutually intelligible among our ancestors. This may have in turn sealed a language’s “first words” and passed on to the successors, before it eventually developed along all manner of creative trajectories.