The Watcher

In the quiet central Texas city of Bryan, elderly retiree George Demoupoulous dreamed of insects. For twenty years the brilliant but increasingly delusional old man had been trying to talk to them.

George was possessed of the bizarre but unshakable belief that, given the enormous number of insect species in the world, surely one of them must be able to communicate with humans. A retired engineer, George had worked out the sorts of signals insects could perceive. He then used this knowledge to build devices that insects could use to send signals back to him. 

Brilliant as he was, George’s project should have been doomed to failure. His devices could have plausibly worked given the right conversation partner, but none of the thousands of insect species in the world had even the most basic capacity for language.

George was undeterred. He just needed to keep spending what remained of his pension on more mail-order insect eggs, long road trips, and flights to exotic locations. Eventually, he would find the right insect. And just as a backup, he kept a few of his prototype communicator devices plugged into each of the ant colonies in his backyard, sending a default message on a continuous repeating loop.

George had no living family that he knew of, and no close personal friends. Therefore, the only potential witness to his goings on was one Jeff Boudreaux, a graduate student renting out the accessory dwelling unit on George’s property. Jeff was a reliable tenant who mostly kept to himself and spent most of his waking hours toiling away in a windowless university basement.

Thus it was that no one had seen George working long hours into the night, day after day. No one had read his lengthy instruction manuals for operating the communicator. No one had heard him talking to himself for hours on end.

No one, that is, except the Watcher.

No human had ever seen the Watcher–because it cannot be seen–and no human had ever heard the Watcher–for it cannot be heard. Yet from the very dawn of mankind, the Watcher had stood patiently by, watching in silence, waiting for the appointed time.

Now.

The Watcher carefully observed a hollow log, home to a colony of carpenter ants. Then, for the first time since it appeared on Earth ages ago, the Watcher did more than just watch. It stretched out its hand and touched the tiny creatures.

Its long mission finally complete, the Watcher departed, never to return.

The next morning, the ants started talking to George.

First Contact

George’s ant communicators were made from off-the-shelf electronics and cost about fifty dollars in materials, consisting of two input/output “terminals” connected by a series of carefully insulated (and meticulously ant-proofed) cables. The terminal on his side consisted of a simple box featuring a black-and-white display panel and keyboard, while the terminal on the ants’ side consisted of a set of audio speakers and an array of buttons specially designed to be operated by ants.

George could type a series of symbols into his terminal and then press “send,” which would cause a corresponding series of sounds (all outside the human range of hearing) to play on the ants’ side. The ants in turn could trigger any of the device’s buttons from their end, which would do two things at once: play a sound for the ants to hear (so they could associate each discrete input with a particular sound), and also send a signal to the other terminal, displaying the symbol corresponding to that sound on George’s screen. 

Essentially, George could send a text message to the ants, which the ants would perceive as “speech” on their side through the vibrational music of the speakers. Then, the ants could “speak” back to him by triggering the inputs, crafting a message from the same set of sounds they were hearing, which the device would render on George’s display as text.

In George’s mind this was all that was necessary, in principle, to establish communication with a species with the sufficient capacity for language. 

The hard part, of course, was finding such a species.

The Watcher had solved that part of the problem, though neither George nor any other soul would ever discover the specific cause. Oblivious to last night’s silent miracle, he stepped into his workshop.

George nearly spit out his coffee. Communicator A3’s notification light was blinking furiously at him. A message was waiting. A real message. 

It was essential to differentiate a true message from random noise, because this wasn’t the first time he had gotten a signal from this device. This was because ants were always accidentally bumping into the communicator buttons every now and then out of chance, with no meaningful pattern behind the signals. George had studied such replies in earnest early on, decided they meant nothing, and then programmed the device to filter them out as he awaited an unambiguous, deliberate, signal. That was the kind he was now receiving, the first since he had set up the background noise filters.

He swelled with anticipation, and eagerly pressed the “messages” button. A string of characters leaped onto the screen. His excitement turned to confusion as he recognized the text – it was the same message he had been broadcasting on loop for months on end. Suspecting electronic interference or some other such glitch, he picked up his screwdriver and started to disassemble the communicator. Suddenly George slapped his forehead in realization. Of course!

Could the ants be repeating his own message back? Wasn’t that a common first step between two unknown parties establishing a communication protocol from scratch? Repeating back the signal could amount to, “Hello, we heard you.” He turned off the broadcast loop and typed a brand new reply, this time using a unique sequence of symbols. They had no meaning just yet, of course, but that would soon change. For now George wanted to see if the ants would repeat this message back to him, too.

Nothing happened.

He sent the message again.

Still nothing.

George was patient. He pulled out a mystery novel and sipped his coffee. Some time later, the communicator blinked again. 

There was a message. Its text was identical to the new sequence George had just sent.

George tried another test. The previous messages had only been five characters long. This time he typed a sequence twenty characters long. When he returned from lunch, he happily noted a third message, the exact same twenty characters, all repeated back in proper sequence.

Don’t be too hasty, George thought to himself. It could still be a glitch in the electronics, the input buffer leaking into the output buffer. But then why the delay?

Fortunately, George had a special debugging tool devised for just this situation. He fumbled around for the remote control and turned on a small CCTV monitor, setting it to camera A3. The black and white screen showed a night-vision display of the interior of the carpenter ant nest, the feed came from a tiny camera mounted directly on the ant-side terminal. He set the playback head’s timestamp to just before the time the latest message was received, and hit play. George dropped his coffee.

Before each of the input triggers stood a single ant. One stepped up to its trigger and activated it. Then another, and another, each in turn, in careful sequence. The inputs matched the symbols George was receiving exactly. 

This wasn’t a bug in the electronics, this was a real message. A deliberate one. This was the moment George had been waiting for all his life. First contact with an alien species, and without even having to leave planet Earth!

Unfortunately, George couldn’t talk to the ants just yet. Sure, he could send them messages, and they could send messages back, but at present the messages were equally meaningless to both parties, essentially just gibberish.

Bootstrapping a language from scratch isn’t easy. Even if two parties have established a channel to talk to each other over, one can’t just start with “Hello, what is your name?” That’s because language is a set of shared symbols whose meaning is agreed upon by its speakers, and at the beginning, the two parties don’t agree on anything, let alone which words mean different things.

This is further complicated when one can’t even assume that the other party shares certain fundamental concepts with you – concepts like “is”, and “name”, and “you.”

This did not bother George, because George was patient. He had plenty of time to work out this next problem. At that thought, his attention fell on his weathered hands at the keyboard, skin loose and thin, veins and knuckles in sharp relief. Well, maybe not plenty of time. But enough.

Communicating with a species so different than himself presented other issues. If he had been sitting across the table from a humanoid space alien out of the movies, he could start by pointing to a cup and saying the word ‘cup.’ Then the alien would point to it and make some utterance in his own language, and they would thereby slowly build up a shared vocabulary of basic nouns and verbs. 

George couldn’t do this with the ants, because they couldn’t even see him. Ants have eyes of course, but their vision is poor even for insects, and many species of ants are even blind. Even supposing that the ants could see him, his gestures would still be meaningless and alien given their vast differences in size and physiology.

Unless he could reach down into their world and meet them on their level, there would be no way to ground this new language in concrete meanings tied to the real world.

George went out to the yard and inspected the buried cable connecting it to the communicator. I need to watch them while I talk to them, he thought. He went back to the garage and got a trowel, his toolbox, and a lawn chair. He spent a few minutes digging up the cable, spliced in a series of connectors, then re-buried it, leaving the branching connector protruding above ground. He then returned to the garage, picked up a spare terminal, carried it out to the lawn with him, and attached it to the exposed connection. He sat down in the lawn chair and held it in his lap, eyes fixed on the nearby ant colony. Now what? He thought.

Ants were curious, and one of their chief perceptions was smell. He went inside and returned with a sealed Tupperware container. Let’s try some experiments. 

He opened the Tupperware container, took out a few grains of rice, and placed them near the nest. He typed R and hit send. Rice. He repeated this message several times. Eventually an ant came, noticed the rice, and went back to the nest. Soon several ants appeared and started gathering the rice, which they quickly took back to the nest. More ants appeared after that to search for rice, but, finding nothing, returned. 

So far this was all typical ant behavior–a scout who finds a source of food leaves a pheromone signal at the location, and then goes back to alert the nest. Other ants follow the scout’s pheromone trail back to the food source, and with each successful new encounter with food, the pheromone trail is reinforced. Once the food is exhausted the next waves of foragers coming to the site find it empty and return empty-handed. At this point they stop emitting pheromones that signal an active food source. As the old pheromones gradually dissipate, the exhausted food source location is forgotten and the scouts and foragers look elsewhere. Bigger food sources that take longer to deplete will result in more ants leaving stronger pheromone trails, and the inverse with smaller food sources. Despite lacking any central authority, in this way the ants naturally allocate more workers to bigger prizes in a dynamic and efficient manner that updates in real time. 

This exact process played out with George’s rice grains. A few ants came to pick up the grains, and several more followed after them, following the trail. But given how small the food source was, it didn’t take long for the ants to stop coming back after all the grains had been collected.

This was what George had been waiting for. So far, the ants probably had no idea what the mysterious messages meant. He placed some more grains of rice and once again sent R. It didn’t take long before a scout found the rice, and foragers soon came and collected their bounty. He offered rice several more times, each time pairing the placements with the message R, the “word,” that, in his mind, meant Rice. 

Finally, he broke the pattern by sending the message without placing any rice, and waited.

Ants came and went looking for something (rice, presumably) and not finding hit.

Eventually, George got a message on his terminal. The display read R. Then again, R. 

He placed another three grains, and then paused at a sudden thought as his fingers hovered over the keyboard. He typed R R R, the same symbol, three times, with brief pauses in between, then hit send. The text spacing was significant and corresponded on the ant’s side to an evenly spaced time delay between the vibrational pulses emitted by the speakers.

Another scout eventually found his three grains, and returned to the nest. This time he noticed a subtle difference in the foraging pattern. Usually after a scout found a food source, the foragers summoned afterwards would mindlessly follow the pheromone trail until it evaporated. This meant that for a typical food source, the colony always inevitably sent more workers than was strictly needed to collect it, because the pheromone trail took time to evaporate even after the last grain of food had been claimed. But this time, George noticed that three – exactly three – workers were sent to collect his three grains of rice. The ants returned to the colony. No other workers visited the now-empty food location. 

Some time later his communicator beeped. The message read R R R. Three-rice. Or was it “Three-food” from the ants’ perspective? He couldn’t know for sure until he introduced further kinds of food paired with different words.

George decided to try something else. He put down a shallow petri dish and squeezed a little water into it with an eyedropper. Then he sent W. Water. It was a hot day, and central Texas ants were always interested in water. The scouts soon found it, and each time they brought it back to the nest, he repeated the message: W. Water. Soon the ants sent back W and he repeated the exercise. He noted the size of water droplet that a single ant could best manage in one drink, and practiced placing exactly that much with his eyedropper. I’ll call this one "antload," the amount of a substance a single worker ant can carry at a time. He placed four drops of water and messaged W W W W. Water. Four antloads. Sure enough, after the water was discovered, the colony sent exactly four ants. No overshoot. After they returned, a message came back: W W W W. George spent the rest of the day establishing a handful of nouns and numerical adjectives. He also tried a few different kinds of food. The ants didn’t seem to differentiate between rice and birdseed, even when he tried using different symbols for birdseed. To the ants, R might have a meaning as broad as “food” or as narrow as “seed/grain.”

He was, however, easily able to get them to understand S (for sugar) as a distinct symbol. The ants also seemed to pick up on sugar-water, SW, quite readily. These two symbols were not played with any pause between them; instead both symbols were sent together as a unit, meaning their sounds were played together simultaneously on the ants’ side. Stringing individual phonemes together in chord-like fashion seemed to be an effective way to build up compound words.

Establishing a working dictionary was a long, laborious process, and both he and the ants would waste many hours and even days confusing one another over the meanings each side was trying to attach to a given symbol. This was further complicated by the fact that the ants had only a limited sense of “where” George was. 

This is entirely leaving aside that “who” and “what” George was were questions whose answers were presently beyond their ability to fathom. Given this tremendous gap between them, it mostly fell to George to introduce new objects into the world the ants could readily perceive, and then guide the negotiations about what to call these things. Fortunately, the ants proved endlessly patient, persistent, and eager conversation partners.

Sometime into the first month, George developed a trick to speed things up by introducing a flat glass plate. Like everything he brought to the ants’ attention, this object unavoidably had his scent on it, and over time both sides had agreed on its symbol. Whenever George wanted to show something new to the ants, he would put it on the plate, then give the word P alongside whatever new symbol he wanted to use. George soon reached an understanding with the ants that when he said, PX, the ants understood that as “Come and see X.”

Soon enough they had a wide variety of nouns for different kinds of food, as well as a generic word: Food. In the end the ants had stuck with R as their preferred word for that; whenever George showed them rice from then on the most specific word they would use for it was just K, the symbol the ants had reserved for Kernel/Grain/Seed.

Nouns came first, and then adjectives. S could be added to a word to approximately mean “sweet.” KS, “Sugar-Seed,” referred to a seed or kernel that was sweeter than average (particularly corn, and especially sweet corn). Likewise WS meant sugar-water. The ants were always interested to know if something had sugar in it.

Verbs came after, slowly and with difficulty. Surprisingly, they were preceded by, and evolved out of, punctuation. This was a multi-step process.

George had started adding a symbol at the end of all of his messages, attempting to indicate a full stop, like in a telegram (This was indicated by ♦ instead of a period as it was easier for his aging eyes to see).

He established this through sheer persistence until the ants picked up on the convention. The word ♦ soon took on a meaning like “now your turn to speak,” which was close enough to what he wanted.

Just shouting nouns back and forth at each other made it a little difficult to understand what the other party wanted of the other. So George got in the habit of teaching the ants to say “please” whenever they seemed to be asking for something. Whenever he would get simple statements of “ R ♦” (Food) from the ants, he would text back two messages: “♥” then after a pause, “R ♥ ♦.”

♥: please

R ♥ ♦: Food, please. Stop.

It took a while but eventually the ants got the message. Every time they wanted something, they would add ♥ to the message. This differentiated a simple statement from a request. And eventually, ♥ started to take on the meaning “want” or “please give.”

Y and N, Yes and No, soon followed, whenever the ants made a reasonable or unreasonable request, and whenever George gave them the right or wrong thing.

However, it soon became clear that ♦ had drifted in the ants’ usage from his intended “telegram full stop” and now just approximately meant “speak” to the ants. He eventually had to establish a new full stop symbol, ◼, and that seemed to stick to its intended meaning.

From these beginnings more verbs came. Want. Give. Take. Go. Come. Search. Find. Then new adjectives. Many. Few. Big. Small. Good. Bad. Tasty. Dry. Wet. Sticky. Dangerous. 

The hardest gap to bridge turned out to be personal subject nouns. I and You and Us were very difficult for George to definitively “point” to. He eventually got there through other means – it didn’t take long to establish a word for ant (A), and eventually they worked out a clear plural inflection that didn’t specify a specific number, just “more than one.” 

“Ants” was thus (A#) – the # being the plural indicator, which played the phoneme with rapid repetitions in a particular staccato pattern. Ants. 

But which ants? There was more than one species, and different colonies in George’s yard and the wider neighborhood. Good ants (GA#) became the closest word George could find in the ants’ vocabulary for “us.” Bad ants (BA#) seemed to mean other colonies.

It also became clear that  A, “ant,” to the ants specifically meant carpenter ants. They utterly rejected any attempt to associate the word George was using for “ant” with any specimen of another ant species, alive or dead. They would either assign it a unique name which he took to be their word for that specific kind of ant, or they would use an oft-repeated word which he took to mean “other/foreigner” and encoded as F, which they also used for other kinds of insects.

But the ants still used the two terms good ants and bad ants to distinguish between different kinds of carpenter ants. George noted that carpenter ant specimens from nests with communicator terminals who had never spoken back to him were consistently identified as bad ants. One day he presented a specimen of bad ants to his colony. Upon receiving the predictable reply, bad ants, he probed further: ♥♦#◼. Please say more. Stop.

The reply came back: 

BA# D◼: Bad ants. Dangerous. Stop.

> Say more.

BA# F GA#◼: Bad ants. Fight good ants. Stop.

> Please say more. Stop.

BA# N♦◼: Bad ants. No speak. Stop.

So… Good ants speak, Bad ants are silent? thought George. Or are they saying they don’t want to talk about it?

George had recently instituted ►, meaning “is a kind of” by putting different pairs of items on the dish at once and carefully repeating it until the ants got the message. K ► R. Seeds are a kind of food. The ants had replied back, Y. Yes, we agree.

Then, a breakthrough. George picked up the plate and put it down in front of the ants. He carefully picked up some good ants on an index card and gently dusted them off onto the plate. He did this repeatedly while sending, “PGA#” over and over. Come see good ants.

The ants seemed a bit perplexed and sent back a new symbol, which George encoded as “?”, hoping it would serve as a useful candidate for “why?” or “am confused” in the future. 

Despite their apparent bewilderment at George’s strange message, some worker ants came to visit the dish and observed that, indeed, some other worker ants from their colony were on the dish. The ants on the dish had previously been scurrying about, but when their compatriots reached them, they all stood still and waited.

The ants messaged back: ♥♦# ♥♦#: Please say more.

George replied: PGA#: Come see good ants.PA♦#: Come see speak-ants.

PGA#: Come see good ants.

PA♦#: Come see speak-ants.

GA# ► A♦#: Good ants are speak-ants.

The ants replied:

Y◼ GA# ► A♦#◼ Y◼: Yes. Good ants are speak-ants. Yes.

Speakers. Little Speakers.

George said:

GA# ► LT♦#◼: Good ants are little things-that-speak. Good ants are little speakers.

The ants agreed:

Y◼ GA# ► LT♦#◼ Y◼

Yes. Good ants are little speakers. Yes.

Then came a surprising message from the ants.

PBT♦◼: Come see big speaker.

Big speaker? He thought. I wonder what that could be.

Again the ants replied: PBT♦◼: Come see big speaker.

Were they about to introduce him to the queen? Or a really large ant? He waited.

Again, and again. came their reply, but no ants came from the nest.

Suddenly he realized what they meant. They were invoking the naming ceremony.

He reached down and pressed the tip of his index finger on the glass plate. One of the ants came up and brushed it with its antenna, then returned to the others who started moving in a flurry of excitement. About half the ants traveled back to the nest. Soon the reply came back.

BT♦◼ YN◼: Big speaker. Yes-no.

He replied back:

Y◼ BT♦◼ Y◼: Yes. Big speaker. Yes.

They are the Little Speakers.

I am the Big Speaker. 

From here a whole slew of concepts were unlocked. With a durable way to refer to each other, they soon worked out something like “You” and “I” and “We” – though the way this worked out “I” and “You” had the underlying literal meanings of “that which is speaking now” and “You” actually meant “that which is being spoken to.”

George noted that the Ants always attached the plural affix to their use of “I” – making it “We,” but always referred to George in the singular. And having unlocked enough basic building blocks, a whole host of complex communication followed.

Finally, after six months of hard work, George could communicate in entire sentences with the ants. There were limitations to be sure, but it was real, genuine language. He could understand the ants. And they could understand him. He was the first speaker of a full-fledged human/ant pidgin, which he dubbed “Humantish.”

George spent the next two months documenting everything he had achieved. As he sat at his workbench, he marveled at what he had accomplished, distilled into his two most precious possessions: firstly the communicator, and secondly its instruction manual: How to Talk to Ants, by George Demoupoulous, First Edition.

What’s next? He thought to himself. Publish. I should get someone to publish this. George had been so single-minded in his pursuit that he had nearly forgotten how to relate to the wider world about him. Tomorrow morning, I’ll ask Jeff to put me in touch with someone at the University. I’ll publish my manuscript. I’ll win the Nobel Prize–maybe more than one! I knew it. I always knew it. And now the whole world will know too.

His heart leapt at the thought and his pulse quickened as he imagined humanity’s bright future alongside the ants. Together, there’s nothing we can’t do. His heart beat faster. 

A smile crossed his face as a massive heart attack seized his elderly body. He was still smiling as he collapsed on the floor of his workshop.

To be continued...