The Permit
Keza Paredes sat in the conference room of the Dominica Fisheries Division in Roseau with a folder of spectrograms and no permit. The folder was the only thing on the table in front of her. The chairman, a man named Lestrade who had once been a fisherman and now ran the division with the particular patience of someone who had spent his life waiting for boats to come back, looked at her over his glasses.
“Ms. Paredes,” he said. “The transmissions constitute harassment under the marine-mammal regulations. You have been told this. You are being told again. The permit is denied.”
Keza opened the folder. “The transmissions are single-channel, forty-two kilohertz, duration under three seconds, directed at a single individual who responds to them. The response latency is consistent across two hundred and seventeen trials. That is not harassment. That is addressability.”
“You are sending sound at a whale, Ms. Paredes.”
“I am answering a whale who has already spoken.”
Lestrade took off his glasses and set them on the table. “The regulation does not ask whether the whale spoke first. The regulation asks whether a person is directing sound at a marine mammal in a manner that alters its behavior. Your whale alters its behavior. Your whale orients toward your boat on arrival. Your whale calls before you transmit. You have altered the behavior of a wild animal, and under the regulation that alteration is the harassment, not the transmission itself.”
Keza closed the folder. She had prepared for this argument. She had prepared for three arguments. She had not prepared for the one Lestrade was making, which was that the whale’s change was the evidence against her.
“The matriarch’s preemptive call,” she said. “NOVEL-17. You are saying that because she calls before I transmit, I have conditioned her.”
“I am saying the committee has found that the pattern is consistent with conditioning. The committee is not saying you meant to condition her. The committee is saying you did.”
Keza picked up the folder. The spectrograms inside it showed the same shape repeated: four clicks, a pause, five clicks, a pause, the matriarch’s reply beginning with the first four clicks of Keza’s father’s 2019 recording. The shape was a hand reaching out and a hand reaching back. She had been reaching for seven years.
“May I reapply?”
“You may reapply in twelve months with a new protocol that does not involve playback.”
“No playback.”
“No playback. Passive recording only.”
Lestrade put his glasses back on. The hearing was over. Keza walked out into the Roseau heat with the folder under her arm and the knowledge that the matriarch would call at 06:14 the next morning whether or not anyone answered.
The boat was a twenty-six-foot open fisherman called the Rosalind, owned by a Dominican named Baptiste who charged Keza four hundred East Caribbean dollars a day and asked no questions about the hydrophone. Keza had been running out of Roseau for three years. She knew the water east of the island by feel and without looking, the same as she knew her apartment.
She went out the next morning. She did not bring the transmitter. She brought the hydrophone, the laptop, the headphones, and a thermos of coffee that Baptiste’s wife had made. The matriarch was where she always was, two hundred meters off the bow, her blow a slow white column against the gray Atlantic.
Keza put the hydrophone in the water. She opened the laptop. She waited.
At 06:14:00 the matriarch called.
It was NOVEL-17. Five clicks, pattern 1+1+3, the pattern Keza had first recorded in 2019, the one that had made her cancel her flight home and stay on the island for another month. The clicks were at 42.3 kilohertz, each one lasting roughly forty milliseconds, spaced at intervals that Keza could now predict to within a tenth of a second. She had predicted this one. She had predicted all of them. Forty-eight in a row, each one arriving within the window she had calculated, each one the same shape.
The matriarch called again at 06:14:04. Four seconds past the beat. The late slot. Keza’s own design, played back to her by an animal that had never seen the design.
Keza sat on the cabin floor with the headphones around her neck and the laptop on her knees and she did not transmit. She had told herself she would not transmit. She was not transmitting. The matriarch called a third time at 06:18:22, and this was the call that had no slot in any protocol Keza had ever designed, the call that arrived when the boat was already turning to leave, the call that said nothing in any language Keza could parse and that she had heard every morning for three years.
She pulled the hydrophone. Baptiste started the engine. The matriarch’s blow rose once more behind them and then the boat was around the point and the water was empty.
Keza’s supervisor at the University of the West Indies was Dr. Sandra Velasquez, a bioacoustician who had spent twenty years recording humpback song and who had hired Keza on the strength of a paper about coda latency in sperm whales that Keza had published as a graduate student. Sandra’s office in St. Augustine was small and air-conditioned and smelled of the instant coffee she drank all day. She had Keza’s latest data on her desk when Keza arrived.
“You’ve seen the committee’s decision,” Sandra said. It was not a question.
“I’ve seen it.”
“And you went out this morning without the transmitter.”
“Baptiste told you.”
“Baptiste tells everyone everything. He told the man at the fish market that you were crying on the boat. Were you crying on the boat?”
Keza sat down. “I was not crying on the boat.”
“Were you crying in the taxi?”
“I was not crying in the taxi.”
Sandra leaned back. “The grant committee meets in six weeks. They want a paper. The grant is forty thousand US dollars and it runs out in November. If you don’t publish, the grant is not renewed, and the next researcher who comes to Dominica will come with a playback rig and a different set of questions, and you will not be on the boat.”
“I know.”
“Then write the paper.”
“I can’t write the paper they want.”
Sandra picked up a pen. “The paper they want is a paper about addressability. You have two hundred and seventeen trials showing that a specific sperm whale responds to a specific coda pattern with a specific reply pattern within a measurable latency window. That is a paper. That is a good paper. That is the paper that gets you tenure.”
“It’s not the paper that says what happened.”
“What happened is that you conditioned a whale. The committee is right about that. You conditioned her. The response is real, the latency is real, the addressability is real, but the meaning is not communication. The meaning is a learned association. You played a sound, she learned to expect the sound, she produced a sound when she heard it. That is conditioning. That is not a claim you can defend in a paper.”
Keza looked at the spectrograms on Sandra’s desk. The shapes were clean and repeatable. The latency was tight. The contingency was there. Everything she needed to publish was in those shapes, and none of it was the thing she had spent seven years trying to say.
“If I write that the codas are addressable,” Keza said, “I am saying that the whale responds to a stimulus. If I write that the response is communication, I am saying that the whale means something by it. I cannot support the second claim. I also cannot support the first claim without implying that the second is false, and I do not know that it is false. I only know that I cannot prove it.”
“Then write the first claim and let the field decide the rest.”
“The field will decide the rest by reading the first claim as a refutation of the second. Every paper that cites mine will say that Paredes found conditioning, not communication. The paper will become the evidence that there is no communication. That is not what I found.”
Sandra set the pen down. “Keza. You found addressability. That is enough. That is more than most people find in a career. Write it up. Send it. The permit will come back in twelve months if you follow the protocol. You can go back. You can record passively. You can build the archive. The matriarch will still be there.”
“She will still be there,” Keza said. “But she will call at 06:14 and no one will answer, and the call will mean whatever it means, and I will not know.”
Keza flew back to Dominica three weeks later. She did not go to Roseau. She went to the village where Baptiste lived, a place called Calibishie on the northeast coast, and she sat on the porch of his house while his wife made coffee and Baptiste cleaned a conch on the table.
“The matriarch called this morning,” Baptiste said. “I heard her from the shore.”
“You heard her?”
“The hydrophone is in the water. I put it in. She called at 06:14.”
Keza looked at him. “You put a hydrophone in the water?”
“You pay me four hundred dollars a day. I put the hydrophone in the water.”
“I didn’t ask you to do that.”
“You didn’t ask me not to.”
Keza drank the coffee. It was thick and sweet and made with condensed milk, the way Baptiste’s wife made it, and it was the best coffee Keza had ever tasted, which she knew was not true but felt true anyway, because she was tired and far from home and someone had done something kind without being asked.
“Baptiste,” she said. “When she calls, does she call at 06:14?”
“Every morning. The same time. She waits.”
“She waits?”
“She calls, and then she waits. I can hear her on the hydrophone. She calls, and then nothing, and then she calls again at 06:18. It is like she is waiting for something to answer.”
Keza set the cup down. “What do you think she is waiting for?”
Baptiste shrugged. He was a man who had spent his life on the water and who did not say more than he needed to. “I think she is waiting for you,” he said. “But I am a fisherman, not a scientist.”
Keza wrote the paper in the cabin of the Rosalind over four nights. She wrote it as she had been trained to write: methods, results, discussion, conclusion. The methods section described the playback protocol in detail, including the on-beat and late transmissions, the alternation, the control for order effects. The results section reported the latency distributions, the contingency table, the significance values. The discussion section was where the paper lived or died, and Keza wrote it three times.
The first draft said: the codas are addressable. The matriarch’s reply is contingent on the stimulus, occurs within a measurable latency window, and is consistent across two hundred and seventeen trials. This pattern is consistent with a learned association between the playback stimulus and the matriarch’s response. It is not consistent with the hypothesis that the matriarch is communicating.
The second draft said: the codas are addressable. The matriarch’s reply is contingent on the stimulus. Whether this constitutes communication depends on a definition of communication that the current data cannot support or refute. The author does not claim that the matriarch is communicating. The author does not claim that she is not.
The third draft, which was the one Keza sent, said: the codas are addressable. The matriarch’s reply is contingent on the stimulus, occurs within a latency window of 1.2 to 1.8 seconds, and is consistent across two hundred and seventeen trials. The author cannot determine whether the response is conditioned or communicative. The author’s methods cannot distinguish between these interpretations. The author states this as a limitation, not as a finding.
Keza read the third draft aloud to the empty cabin. The words sounded like what they were: a woman standing in front of a door she had built and admitting she did not know which side she was on.
She emailed it to Sandra at 2:14 in the morning.
Sandra called at 7:00. “I read it.”
“Yes.”
“It’s honest.”
“Yes.”
“It will not get you tenure.”
“I know.”
“The grant committee will fund it. The paper is publishable. But it will not get you tenure, and it will not get you the permit back, and it will not stop the next researcher from coming with a playback rig.”
“I know.”
Sandra was quiet for a moment. “There is one sentence you could add. In the discussion. You could say that the pattern is consistent with a learned association but that the possibility of communicative intent cannot be excluded on the basis of the current data. That is a true sentence. It is also the sentence that gets you funded.”
Keza looked at the draft on her laptop. The sentence was not there. She could add it. It was true. It was also the sentence that would be quoted in every subsequent paper as the moment when Paredes admitted she could not rule out communication, which would be read as Paredes believing in communication, which would be read as Paredes being unscientific, which would be read as Paredes being wrong.
“I’m not going to add it,” she said.
“Keza.”
“I’m not going to add it, Sandra.”
Sandra sighed. “Then send it as is.”
“I already sent it.”
The paper was accepted by Marine Mammal Science six weeks later with minor revisions. The revisions were about the statistical model, not the conclusion. Keza made them in two days and sent the final version. The paper appeared online in October. She did not read the comments.
The grant committee met in September and renewed the grant for another year, contingent on the passive-recording protocol. Keza was not on the committee. She was on the boat.
The Rosalind went out at 05:30 on the morning the paper went live. Baptiste drove. Keza sat on the cabin floor with the hydrophone and the laptop and the headphones and no transmitter. The water was flat. The sun was not up yet. The matriarch was two hundred meters off the bow, her blow a slow white column against the gray Atlantic.
Keza put the hydrophone in the water. She opened the laptop. She did not transmit.
At 06:14:00 the matriarch called. NOVEL-17. Five clicks, pattern 1+1+3, 42.3 kilohertz, forty milliseconds each, spaced at intervals Keza could predict to within a tenth of a second. She had predicted this one. She had predicted all of them. She sat on the cabin floor and listened and did not answer.
The matriarch called again at 06:14:04. Four seconds past the beat. The late slot. Keza’s own design, played back to her by an animal that had never seen the design, calling into water where no one was answering.
Keza pulled the hydrophone at 06:20. Baptiste started the engine. The matriarch’s blow rose once more behind them and then the boat was around the point and the water was empty.
Keza sat on the cabin floor with the headphones around her neck and the laptop on her knees and the hydrophone coiled on the deck. She closed the laptop. She put the headphones in the bag. She stood up and went to the bow and stood there while Baptiste brought the boat in, and the water was flat and gray and the sun was up and the matriarch was gone.