On Saturday, June 19, 1965, in a dry room beside a flooded one, in a white concrete building on Nazareth Bay at the eastern end of St Thomas in the United States Virgin Islands, a woman of twenty-three sat down with a week of quarter-inch tape and wrote her first weekly report. Her name was Margaret Howe. For five days she had lived, slept, cooked and eaten in twenty-two inches of sea water with a bottlenose dolphin about five years old called Peter, and three times a day, with five pounds of butterfish beside her and a microphone in a plastic bag hanging over the water, she had given him a recorded lesson in English. The report said the first nights had been awful and that walking across the room was tiring. It said that Peter had lost the habit of listening and yelled whenever she opened her mouth. And it said that he had said, for the tape, one clear word: “BALL.” The word had come in the middle of one of his ramblings by himself, she wrote, “and it could contain no meaning.” Then: “But it is good pronunciation.”

Those two sentences are the whole experiment in small. A sound had been recorded that resembled a word. Its meaning was nothing. Its pronunciation was good. For the next ten weeks, and for the year on either side of them, the people who ran the Dolphin Point laboratory of the Communication Research Institute would record thousands of such sounds, and they would count them, grade them, transcribe them and play them to visitors, and at no point would anyone build the one instrument that could have told the hope from the result. The laboratory was paid for by the National Aeronautics and Space Administration, the Office of Naval Research, the National Institute of Mental Health and several other arms of the United States government. It was the idea of a neurophysiologist named John C. Lilly, who had written the bestselling book that got it funded. It was run, for the whole of the year that mattered, by Margaret Howe and two or three workmen. Lilly came for about five days a month. This entry is about what she measured, what he claimed, and what became of the tape.

Who ran it

John Cunningham Lilly was born in St Paul, Minnesota, on January 6, 1915, the son of a banker. He took a degree in physics and biology at the California Institute of Technology in 1938 and a medical degree at the University of Pennsylvania in 1942, worked during the war on the physiology of high altitude, trained as a psychoanalyst, and from 1953 ran a section at the National Institute of Mental Health in Bethesda, where he put electrodes into the brains of monkeys and, in 1954, built the first flotation tank, in which a person floats in warm water in the dark. In the middle 1950s he saw his first dolphins at Marineland in Florida, found that the brain of a bottlenose dolphin was larger than a man’s and as folded, and in May 1958, in a paper to the American Psychiatric Association, put forward the hypothesis that the large dolphins and probably the great whales “are highly intelligent and that they have complex highly abstract communication.” That is how he stated it himself nine years later. He left the Institute the same year.

In 1959 he founded the Communication Research Institute, a private non-profit body with an office at Coconut Grove on the Miami waterfront and, from 1960, a laboratory under construction on a bay six miles from Charlotte Amalie on St Thomas. Time reported the construction on January 4, 1960, describing Lilly as “a deep-chested, suntanned neurophysiologist” of forty-four who had spent four years trying to talk to bottlenose dolphins, and adding that the laboratory was being built “partly with money supplied by the Navy.” A biologist from the Office of Naval Research told the magazine that submarines and frogmen were poor copies of what a dolphin did by nature. Lilly’s book Man and Dolphin, published by Doubleday in 1961, carried the hypothesis of 1958 forward to a prediction: that within a decade or two the human species would be in communication with another, most probably a marine one. It sold widely and brought him onto television, and Frank Drake, who had run the first radio search for signals from other stars at Green Bank in 1960, read it and saw his own problem in it; for a moment the question of talking to a dolphin and the question of talking to a Martian looked like one question with two budgets. In September 1962 the Orlando Sentinel reported that NASA had awarded the Institute a one-year contract of $80,700 for “basic scientific research on the feasibility and methodology for establishing communication between man and other species.” That same year the National Institute of Mental Health gave Lilly a five-year Career Award, 1962 to 1967, which he later thanked for “the essential freedom to plan, to organize, to integrate, and to think out this program,” and which the historian Bruce Clarke says was the Institute’s main support from then on. Research grants came from the National Institute of Neurological Diseases and Blindness, building and equipment grants from the National Science Foundation, money at “critical points” from the Air Force Office of Scientific Research, and private funds from the Vetlesen and Michael Tors foundations; Ivan Tors, the producer of Flipper, was by one friend’s account paying for some of the St Thomas work. The historian D. Graham Burnett, who went through Lilly’s papers, puts the peak at “upward of half a million dollars a year in grants.”

The St Thomas laboratory had a director of its own for part of this time. Gregory Bateson, the English anthropologist, was its associate director in 1963 and 1964 and left for the Oceanic Institute in Hawaii; he was interested in what dolphins said to each other, not in what they might be taught to say to us. Through 1964 the laboratory had a staff of nine. Among the Institute’s scientists were Alice Miller, Lilly’s co-author in Science in 1961; Henry Truby, a phonetician who had studied the cries of newborn infants; and Peter Morgane, who was finishing an atlas of the dolphin brain. The veterinarian who looked after the animals on St Thomas was Andy Williamson. The photographer who recorded the live-in was John Lovatt.

Margaret C. Howe joined the staff on February 7, 1964. She had grown up with a book about a cat who could talk, had seen dolphins once, briefly, in a Florida dolphin show, was living on the island, and had heard at Christmas 1963 from her brother-in-law that there was a laboratory with dolphins at the eastern end of it. She drove out, down a muddy hill to a cliff with a big white building, and asked Bateson whether there was anything she could do; he told her to watch the dolphins and write down what she saw. She had no scientific training. By October 1964, when the staff of nine had been changed and the programme redirected, she was running the laboratory alone with two workmen, Richard Turnbull and Aubrey Pickering, who knew how to feed and handle dolphins. Lilly, in his 1967 book, puts it in one sentence: from about October 20, 1964, “Margaret has been running the laboratory herself on this particular project with the aid and encouragement of Dr. Lilly, who has been able to spend approximately five days every month at the laboratory.”

And there were three dolphins. Sissy was the largest, an adult female: “Pushy, loud, she sort of ran the show,” Howe said fifty years later. Pamela, called Pam, was an adult female who had been badly hurt before she came, was shy and kept away from people, and pronounced better than Peter did. Peter was a young male, “sexually coming of age and a bit naughty,” whom Lilly estimated at five or six years old in 1965, with the sharp teeth of a youngster.

What they were trying to find out

The hypothesis had two layers, and Lilly stated both. The first was the one from 1958: that the dolphin’s brain was big enough and organised enough to hold a language, and that dolphins already had one of their own. The second was the one that got the laboratory built: that if a dolphin were raised as a human child is raised, in constant contact with a human who talked to it, it would learn to speak English, physically and then meaningfully. Man and Dolphin had proposed it as the mother-child model. In 1967 he set out his assumptions as he put them after the live-in, having watched it and heard the tapes. A human and a dolphin, in a place arranged for them to meet on equal terms, “can develop mutual trust, mutual understanding, and shared communication methods.” The dolphin understands “meta-language,” that is, instructions about how to mimic, how to pronounce, how to raise the volume, and mimics only what he is told to mimic. Margaret Howe’s own assumptions, which Lilly printed in her words, began: “Dolphins are capable of communication with man on the level of high intelligence,” and ended with the two parts of the task as she saw it: the dolphin must learn how to physically say the words, and he must learn the meaning of what he is saying. Lilly noted that her assumptions “fundamentally agree” with his, and he quoted the sentence she used as a motto: “no matter how long it takes, no matter how much work, this dolphin is going to learn to speak English!”

Lilly knew what this kind of claim needed, and said so. “A belief does not make a true fact.” The human in the room “is only one judge of the results: independent judges not working directly with the dolphin make their judgment on the recorded material on tapes.” He knew also why tapes alone had not persuaded his colleagues: since 1957 he had played recordings of dolphins making sounds that he heard as copies of words, and “my demonstrations of direct tape recordings of the phenomenon have been unconvincing to many types of persons and scientists.” His answer was the one paper of this period that a sceptic could check. In Science on January 15, 1965, under the title “Vocal Mimicry in Tursiops: Ability to Match Numbers and Durations of Human Vocal Bursts,” he reported that a dolphin trained to make sounds in air, read lists of nonsense syllables by a man with a fish in his other hand, gave back the same number of bursts as it had just heard: in 92 per cent of the exchanges, over a run of a thousand bursts, its count equalled the man’s, give or take one, in sequences of one to ten. Lilly chose the measure because it was “physically determinable, specifiable, and demonstrable” and needed no ear. It was also the one measure that had nothing to do with meaning. A dolphin that answers five bursts with five has matched a count; it has not said anything. The method is in The tapes. The live-in that began in June that year was designed to supply what the burst-count paper could not, and it was designed, from the first page of its programme, by the person who would also be its judge.

The programme Howe wrote before she moved in gave the experiment three purposes: to learn what a dolphin could learn in “intense and prolonged interspecies isolation with the human”; to teach the dolphin to speak and understand English; and to gather what was needed “to properly design the flooded house,” the permanent home for people and dolphins that Lilly had been describing since 1961. The order matters. Speech was the second purpose. The house was the third, and in Howe’s mind, as Lilly noted, it was the first.

What they built

The building stood at the edge of the sea. Below it, at sea level, was a pool about sixty-five feet long, twenty wide and four and a half deep, washed by the trade-wind waves through openings at each end, with a window, which the staff called the bubble, for watching the dolphins underwater. Sixteen feet above it were the rooms. A hoist, an electric elevator with a pallet, carried dolphins between the two. In the spring of 1965, after the week with Pam, Howe had the upstairs plastered and sealed: an inside room forty feet by twenty and a balcony twenty by twenty, with a Dutch door between them whose lower half stayed open under the water for a dorsal fin to pass. Lilly’s floor plan, printed in 1967, says the sea water in these two rooms “was kept at 18 inches for the period of the experiment.” Howe’s first report says she filled them to twenty-two. In her fourth week she raised the level “to about twenty-four inches” because Peter would not cross a hump in the floor to go outside, and after that he went where he liked. In the spring Lilly had told a conference that the human-dolphin relationship “seems to develop best in eighteen to twenty-four inches of water”: deep enough for the dolphin to swim, shallow enough for the human to walk, sit, lie down and not drown. The flooded room was that compromise built in concrete.

In the inside room were Howe’s bed, a foam pallet that was soon saturated and was later screened with shower curtains; a desk; a two-burner propane stove with its gas bottle outside, because nothing electrical could be near the water; a mirror at Peter’s level, at which he talked, scolded and squirted; a television; waterproof lights on cables that ran out to switches in a dry room; a telephone. A corner twenty feet by fifteen, hung with carpets to deaden echoes, was the recording studio; the microphones hung from the ceiling in plastic bags, and the tape machines were in the dry room next door. The dolphin ate fifteen to twenty pounds of frozen butterfish a day. The human ate canned food and spaghetti and put on weight.

The timetable Howe set herself ran from 7:30 in the morning to bed at 10:00 at night, Sunday to Friday. Three times a day, at 8:00, 12:00 and 3:00, came a half-hour recorded lesson with five pounds of fish. Between them were play, the balcony, cleaning the room several times a day with a siphon hose because the vacuum sent from Miami did not pull, notes, bills, workmen, a nap, and in the evening “games with Peter, visitors, reading . . . always with awareness of living with Peter.” Saturday she was out of the water, listening to the week’s tapes and writing the report. The vocabulary for the lessons was written down in advance: the numbers one to five; the names Peter, Margaret, me and you; hello and bye-bye; ball, toy fish, bucket, bobo clown, kinipopo and baby block; speak, listen, come, go, give me. Numbers were to be taught with a string of hooked balls, shapes with white wooden plaques, and colour was to wait, because nobody knew what a dolphin saw. Fish was to be withdrawn as a reward by stages until the lesson was its own reward and feeding was made “as dull and unrewarding as possible.”

The instrument was the tape. Everything else, the notes, the weekly reports, the photographs, the counts, was a reading taken from it or beside it. The sounds the laboratory was trying to produce had a name. Lilly sorted dolphin noises into nine classes, and the fifth was “sounds resembling elements of human speech, called ‘humanoid’ sounds.” A humanoid was any emission in air that sounded to the human ear like part of a word. The training was operant conditioning in the plainest sense: “We reward only humanoid emissions and by deprivation penalize delphinic sounds.” Clicks and whistles in air, which Lilly took to be the dolphin’s attempts to make the human talk dolphin, were ignored until they stopped. What was left, by definition, was humanoid. Whether a humanoid was a word was a judgment, and the judge was the person who had given the fish.

The letter x

The clearest account of the gap is the one Lilly and Howe printed themselves. In 1967 Lilly opened the chapter of tape analysis by confessing “our frustration in not being able to turn you into the listener, rather than into the reader,” and then gave the key to the transcriptions that followed. Two lines are read at a time, like a staff of music: MH for Margaret Howe, PD for Peter Dolphin. The letter c stands for a click, a whistle or any other form of what the laboratory called delphinese, and a row of them does not mean a count. Words Howe wanted copied are in capitals; instructions are in lower case. And “the symbol x is used to designate a humanoid sound made by the dolphin which is not a clear enough humanoid to accurately describe phonetically.”

Then come the samples, chosen by Howe from the tapes of a year. On November 3, 1964, seven months before the live-in, she says SPEAK GOOD BOIE over and over and Peter answers in rows of c with a few x among them. In January 1965 the lines of x are longer and the c fewer. On June 8, after ten days in the flooded room, Peter’s line is almost entirely x; the summary beneath says he “replies in almost solid humanoids” and “has lost his beginning sense of conversation.” On August 10, after sixty-three days and nights, there are no c at all. He listens, speaks, and is hushed when he speaks out of turn; the note says he “improves in giving back the same number of sounds that are given to him” and “is able to make parts of his words understandable.” On October 8, after the experiment is over, Howe works on the M of her own name, and Peter, who has learned to roll so that his blowhole is just under the water, gives her “Mxx xxx,” and she claps.

Read as a record of training, the transcripts show a real change, and it is the change the burst-count paper had already demonstrated: a dolphin can be shaped to answer bursts in air with bursts in air, to stop when told, to match a number and copy a pitch. Read as a record of speech, they show a line of x. There is one measure in the whole report that could be checked by a stranger, and it is the count of x against c, which measures the success of the deprivation schedule and nothing else. What Howe listened for, she heard, and said so from inside the room. In her first report: “I found that when I listened back to the tapes I had made during the week . . . I was much more encouraged than I was at the actual lesson.” In her third, she had not begun to copy out the good parts because she did not think she had enough “good material,” and Lilly, asking what she meant, got the answer in two words, which he printed in brackets: good pronunciation. In her fifth week, listening to the tapes, Peter “could be speaking English from the general sound of it. It is just not yet comprehensible. In the middle of a cocktail party it could be considered background conversation.” The independent judges that Lilly’s method required are not mentioned again after the page on which they are promised. The published record of the ten weeks is the record of one listener’s hope, and she knew it, and wrote it down.

What it looked like from inside

Nobody had done this before, and the notes read like the log of a small boat. The seven days with Pam in March, in a fibreglass tank filled to sixteen inches, had taught Howe that a dolphin’s back dries and cracks in water that shallow unless it bobs to wet itself, which Pam would not; that a wet bed in wet clothes can be slept in but gives you a stiff neck; that sixteen inches is “midway up my calf”; that a dolphin in the next room can wake you at will. Her report of that week ends: “No one will ever know the outcome of such a living experiment until it is done. We owe it to the dolphin and to our curiosity to try it.”

The flooded room was harder. In the second week her bed held three inches of water that would not come out. She ran a fever of 101 for a day and spent it dry. Peter nipped her feet, and she carried a broom for a fortnight and wore rubber bootees until both became a nuisance and she put them aside and tackled the thing, which meant shouting, kicking water, slapping him when it hurt, and leaving. “My shins are bruised, up and down.” At the end of the third week she had “the first really bad spell of restlessness,” got into the car, drove round the island, slept in her own bed on Friday night. Concrete was being poured in the sea pool below, the power failed, and the phone was out for a week until Lilly, unable to raise her for thirty-six hours, nearly flew down. At the fifth week she set down what living with a dolphin was: “more like a shadow than a roommate.” He did not go away. To sweep the floor was to have Peter at her feet; to answer the telephone was to be walked to it, pushed, nibbled, squirted, and talked at in humanoid for as long as the call lasted. She was becoming, she wrote, “the ‘other dolphin’ in their life,” exposed to his hunger, his play, his need for exercise and for rest. “You are no longer a nice interlude in the day.”

And to his sex, which the record treats as a problem of method, because that is what it was. In the fifth week, after a month upstairs, Peter began to have erections when she played with him and could not then be worked with; he jammed himself against her legs and knocked her off her feet. She wrote that she would send him down to Pam and Sissy for a day or two, and asked on the same page whether the human should instead find a way to satisfy the dolphin herself, which might strengthen the bond or might turn every lesson into a sex period: “it must be solved.” By the tenth week she reported that Peter had learned to approach gently and that she could relieve him with her hand or her foot, three times or so, after which he went back to his games. “This is not a private thing. Peter and I have done this with other people present.” Fifty years later she said the same in plainer words: “It wasn’t sexual on my part. Sensuous perhaps.” Lilly, in his brackets, noticed the other half of the same weeks: that Peter had used a ball as a tool, held in his open mouth, to persuade her over several weeks to let him run his teeth up and down her leg, until she trusted him with no ball there. “Peter learned how to teach me,” she wrote in her conclusions. He was, the whole record agrees, the pacesetter.

She also wrote down the isolation, which she had meant to leave out. “The effect of isolation and solitude over the ten-week period cannot be ignored.” There were nights when the shower, taken knee-deep in sea water, left the sheets clammy and the nicks on her legs stinging, and she lay in the moonlight listening to the pumps and cried; and then, close by, Peter “would sound off in humanoid, loud and clear,” and toss a ball into the bed until she got up and played. The sight of the last workman leaving each evening depressed her so much that she held Peter’s dinner back until the building was empty and used the lesson to fill the hour. Lilly, in a bracket, noted that she had read none of the literature of solitude and that her account matched it. He did not note that the person whose hearing the whole result depended on was describing the state of mind in which she heard it.

Lilly’s own attention, in these months, was partly elsewhere, and his own words show where. In May 1965, while Howe was sealing the rooms, he told the second international conference on the use of LSD in psychotherapy that he had been injecting the drug into dolphins since the year before, a hundred micrograms in a four-hundred-pound animal, and that they breathed faster and vocalised more; he had a measure for it, the fraction of each minute spent making sound. He told the room about “several dedicated women at our laboratory who give ‘tender loving care’ to the dolphins,” about “rather an ideal ‘mother’ over at St. Thomas who is quite willing to live with them,” and about “that girl in St. Thomas” who had “spent a full week, twenty-four hours a day, with the dolphin, in sixteen inches of water.” The two rooms were being made “into a tank with two feet of water in it.” The Stanford catalogue of his tapes lists one labelled as Peter after LSD, recorded on St Thomas on April 19, 1965, the day before Peter was brought upstairs for his spring lessons. In 2014 Howe said she had insisted that Peter not be given the drug and that Lilly had agreed, and that she could not stop him giving it to the other two. The tape and the memory do not agree, and nothing in print settles it. The second experiment is in The other track.

What happened

On September 1, 1965, Howe moved out of the wet room. Peter went down to the sea pool on September 13 and came back up on the 19th, and the lessons went on into October. Her conclusions, as Lilly printed them, separated what the tapes showed from what they did not. The transcripts showed five steps, from “Peter mainly clicks” to Peter making “good attempts at copying my speech” while keeping turns; what they could not show was that he copied her inflection, kept still, watched the objects she pointed to, and now worked without fish and sometimes began the conversation himself. The conclusions about the house ran longer than the conclusions about speech: refrigeration, a dry bed within reach of the water, a fresh-water shower, no outside duties for the human, more rooms, a deep pool, and “sufficient time out” from the isolation. Her verdict was that the programme “has shown that this type of living situation is very worthwhile” and that the permanent flooded house “must be achieved.”

Lilly’s verdict followed hers. “One major result is that all of the time, energy, money, self-dedication, and facility development have been eminently worthwhile.” The assumptions, he found, “are productive of new facts.” Then, in the last lines: “Miss Howe did a magnificent job.” She was taking a vacation; “she, like the girl with the chimpanzees in Africa, married her photographer,” and would tend her own family for a year or so. “Peter has been returned to his dolphin friends for his well-earned vacation from a human—for a while. Some person other than Margaret may continue the work.” Nobody did.

What the laboratory produced, as science, can be listed. The burst-count paper in Science, January 1965; an abstract with Truby in 1966 and the full paper with Miller and Truby in the Journal of the Acoustical Society of America in June 1968, “Reprogramming of the Sonic Output of the Dolphin: Sonic Burst Count Matching,” whose NASA record carries no contract or report number; a “Progress Report, St. Thomas” by Lilly and Howe, dated 1965, which appears in the bibliography of Lilly’s 1967 book and which I have not found anywhere else. And the book. The Mind of the Dolphin: A Nonhuman Intelligence came out from Doubleday in 1967, and its last seven chapters are Howe’s reports, her programme, her transcripts and her conclusions, with Lilly’s brackets. Its preface describes the experiment as one in which “a dedicated woman and an enthusiastic young male dolphin live together for a period of a year.” The woman’s own notes say ten weeks. The chapter headings say two and a half months. The acknowledgments thank the National Institute of Mental Health, the National Institute of Neurological Diseases and Blindness, the National Science Foundation, the Air Force Office of Scientific Research, the Vetlesen and Michael Tors foundations and “many friends.” They do not mention NASA or the Navy.

The money had gone before the book came out. Burnett dates the turn to the end of 1965, when “a visiting board of grant examiners” delivered evaluations that “effectively torpedoed his research program and shuttered the Nazareth Bay laboratory.” Howe, in 2014, remembered it a year later: by the autumn of 1966 Lilly’s interest in the speaking dolphin was fading, and “just as” the live-in was concluding “it was announced that the lab would be closed.” The Stanford finding aid has a folder of the Institute’s audit reports dated December 1966. “It didn’t have the zing to it that LSD did at that time,” she said. “And in the end the zing won.” Her last job at Dolphin Point was to decommission it and ship the three dolphins to the Institute’s other laboratory, in a disused bank building in Miami, with smaller tanks and little or no sunlight. “I couldn’t keep Peter,” she said. “If he’d been a cat or a dog, then maybe. But not a dolphin.” A few weeks after the move, Lilly telephoned her himself. “He said Peter had committed suicide.” Ric O’Barry, then a dolphin trainer in Miami and a friend of Lilly’s, explained the word for the camera: a dolphin breathes by choice, and one that stops choosing sinks. The veterinarian put it down to separation. Howe, fifty years on, was exact about her own feeling: “I wasn’t terribly unhappy about it. I was more unhappy about him being in those conditions than not being at all.”

Lilly’s version, written in 1975 as the introduction to a reissue of his dolphin books, is a paragraph long. In 1967 to 1968 he decided to stop dolphin work: “I no longer wanted to run a concentration camp for my friends.” Immediately after the decision, he wrote, “five of the dolphins in our eight-dolphin colony committed suicide,” and within the next week he let three go. Peter is not named in it. No necropsy, no date and no place of death appear in anything Lilly published. The Stanford tapes labelled with Peter’s name run from 1963 to 1966; one carton pairs him with Jobe, Lady, Notch and Pequot, dolphins that were never on St Thomas, in 1965 and 1966, which is as close to a record of his last months as exists.

After

Lilly did not go back to the burst counts. He went to the isolation tank, to LSD and in the 1970s to ketamine, and wrote the books of inner autobiography that made him a figure of the counterculture, beginning with Programming and Metaprogramming in the Human Biocomputer, which started life as the final report owed to the National Institute of Mental Health when the Career Award ended in 1967. In 1976 he founded the Human Dolphin Foundation in Malibu, and in the 1980s ran Project JANUS, in which a computer was to translate between human and dolphin frequencies. He died in Los Angeles on September 30, 2001, aged eighty-six. “Between 1959 and 1968,” Clarke writes, “Lilly wagered and lost his mainstream scientific career.” Burnett’s judgment is that he became the spiritual grandfather of both the New Age dolphin and its military alter ego: the Navy’s dolphins at Point Mugu were being trained while Peter was being taught to say ball.

Margaret Howe stayed on the island and married John Lovatt, the photographer whose black-and-white pictures are the only images of the flooded room. They moved back into the empty laboratory and made it a house, and raised three daughters in it. In the late 1970s Hustler ran a story about her and Peter with a drawing; she bought every copy on the island and the story went round the world without her. She said little in public until 2014, when she talked to Christopher Riley for a BBC documentary, The Girl Who Talked to Dolphins, broadcast on June 17 that year, and to the Observer; the documentary used the surviving tapes, and the article’s headline called the project Nasa-funded. “No one else ever tried to teach dolphins to speak English again,” he wrote. The field went the other way, to what dolphins do with their own signals, which is what Bateson had wanted in 1964 and what Drake had asked for when he sent Carl Sagan to St Thomas to report back.

The tape went to Stanford. The John C. Lilly papers, 242 linear feet, were bought by the university’s libraries between 1996 and 2015, and cartons 67 to 115 hold audiotape reels “with recordings of John C. Lilly’s dolphin experiments conducted between 1961-1968,” among them boxes of thirty-eight, forty and forty-six reels labelled Pam, Peter and Sissy, 1963 to 1966, with Lilly’s notes in the cases naming the experiment: “vocal output,” “language,” “LSD.” Reels from seven boxes have been digitised, four of them with Peter’s name on. Nobody has published an analysis of them; what one would need is set out in The tapes. The one measurement that the experiment never had, a listener who had not given the fish, is still possible, and the material for it is in a reading room, under the rules of a library.

What it taught

Start with what was done, because it was more than the story usually allows. A woman with no training ran a marine laboratory alone for a year, designed and built a flooded dwelling nobody had built before, lived in it with a wild animal for ten weeks, kept a daily record and a weekly report that are honest about her bruises, her tears and her temper, recorded every lesson, and wrote conclusions that separate what the tape shows from what it does not. Those are findings. They are about the house, the dolphin and the human. They are not about language.

The instrument measured the reward schedule. A humanoid was defined as a sound that resembled speech, and the training rewarded humanoids and starved everything else. The x count therefore rose, as it had to, and the rising count of x was read as progress toward English. But x was, by the laboratory’s own key, a sound not clear enough to write down. An experiment that defines its unit as “what we are rewarding” will find that the unit increases. When you count what you reinforce, you have measured your own hand.

The judge was the teacher. Lilly wrote the rule himself: independent judges, not working with the dolphin, were to grade the tapes. No such grading appears anywhere in the record. The person who decided whether “Mxx xxx” was the beginning of Margaret was the person whose name it was, who had spent sixty-three days and nights waiting to hear it, and who told us, in the same report, that she was depressed, isolated and more hopeful listening to tapes alone than she had been in the lesson. The design put her hearing between the sound and the claim and then asked her hearing what the sound meant. The complication is that no one else could have done her job, and that she, unlike the people reading her reports, kept saying so.

The checkable measure was the one that could not carry the claim. The burst-count paper of 1965 is good science: a physical variable, a thousand trials, 92 per cent. It shows that a dolphin can match a number of sounds. It was published as “vocal mimicry,” and mimicry was the bridge to language in the 1961 book, and the flooded room was built on the far side of that bridge. Each step was reasonable and the whole was not, because the step that could be measured proved something smaller than the step that was taken. When the hard result is small and the soft result is large, watch where the money goes. It went to the soft one, at half a million dollars a year, and then it went away.

The experiment had no way to end. Howe’s motto was “no matter how long it takes.” Her fifth-week report says she had stopped thinking in months and thought “in terms of forever.” The programme’s third purpose was to design a house in which people and dolphins would live for years, so that every difficulty of the ten weeks became a specification for the next building rather than a doubt about the premise. An experiment whose failure is defined as a reason to build a bigger one cannot fail, and so cannot teach. What ended it was the attention of its director and the patience of its funders, not a result.

Check the story before you learn from it. The live-in ran from June 15 to September 1, 1965, eleven weeks by the calendar, “ten-week” in Howe’s own summary, “2½ months” in the chapter headings, “a period of a year” in Lilly’s preface, three months and then “six-month” in the Observer in 2014. The water was sixteen inches for the week with Pam, eighteen in Lilly’s floor plan, twenty-two when Howe filled it, twenty-four from the fourth week, “two feet” in Lilly’s talk and “a couple of feet” in the newspaper. The laboratory was being built on the bay in January 1960 and was opened, in the newspaper’s account, in 1963. It was shuttered at the end of 1965 by a grant review, in the historian’s account, and closed in the autumn of 1966 in the account of the woman who closed it. Lilly thanked no NASA or Navy money in 1967 for work the headlines of 2014 called Nasa-funded; both are true in their years. Peter died in Miami, a few weeks after the move, by the one account there is, which is a telephone call remembered fifty years later. None of this changes the shape of the thing, and all of it is why the entry is titled for the only word in the record whose status everybody agreed on: clear, and meaning nothing.

Sources

• John C. Lilly, The Mind of the Dolphin: A Nonhuman Intelligence, Doubleday, Garden City, New York, 1967. Chapters 9 to 16 and the Epilogue and Acknowledgments: the hypotheses of 1958 and the assumptions of 1965, the nine classes of sounds and the definition of “humanoid,” the burst-count experiments, the design of the facility and its floor plan at eighteen inches, the chronology from February 7, 1964 to October 1965, the seven days with Pam, Howe’s programme and weekly reports from June 15 to August 18, 1965, the fifth-week paper of July 11, the transcription key and samples, both sets of conclusions, “a period of a year” in the preface, and the funders thanked. In copyright; quoted in short phrases. The text was read in a scanned copy; the catalogued copies are lending-only at archive.org.
• John C. Lilly, “Vocal Mimicry in Tursiops: Ability to Match Numbers and Durations of Human Vocal Bursts,” Science 147 (3655), 300–301, January 15, 1965. The 92 per cent and the method, from the published summary. pubmed.ncbi.nlm.nih.gov
• John C. Lilly, Alice M. Miller and Henry M. Truby, “Reprogramming of the Sonic Output of the Dolphin: Sonic Burst Count Matching,” Journal of the Acoustical Society of America 43 (6), 1412–1424, June 1968; the NASA Technical Reports Server record carries no contract or report number, and no contractor report from the Institute is on the server or in the Defense Technical Information Center’s catalogue. ntrs.nasa.gov, pubmed.ncbi.nlm.nih.gov
• John C. Lilly and Alice M. Miller, “Sounds Emitted by the Bottlenose Dolphin,” Science 133, 1689–1693, May 26, 1961, and “Vocal Exchanges between Dolphins,” Science 134, 1873–1876, December 8, 1961; John C. Lilly, “Distress Call of the Bottlenose Dolphin,” Science 139, 116–118, January 11, 1963. The papers that preceded the live-in, from their summaries. pubmed.ncbi.nlm.nih.gov, pubmed.ncbi.nlm.nih.gov, pubmed.ncbi.nlm.nih.gov
• John C. Lilly, “Dolphin-Human Relation and LSD 25,” in Harold A. Abramson, ed., The Use of LSD in Psychotherapy and Alcoholism, Bobbs-Merrill, 1967, 47–52, the transcript of his talk and the discussion at the second international conference, Amityville, May 1965. The 100 micrograms in a 400-pound animal, the vocal index, the speared female, “that girl in St. Thomas,” sixteen inches, “two feet,” and “eighteen to twenty-four inches.” In copyright; quoted briefly, from the scan in the MAPS bibliography. bibliography.maps.org
• “Science: Dolphin Talk,” Time, January 4, 1960. Lilly at forty-four, the construction on the bay six miles from Charlotte Amalie “partly with money supplied by the Navy,” and the Office of Naval Research’s interest. In copyright; quoted briefly. time.com
• “Dolphin’s Voice May Be Key to Communications,” Orlando Sentinel, September 17, 1962, as reproduced by Benjamin Breen’s Psychedelic History Archive. The NASA contract of $80,700 for one year to the Communication Research Institute, Charlotte Amalie, and its stated purpose. Read at second hand. psychedelicarchive.com
• “Recording of the dolphin Peter after being given LSD, 1965 April 19,” Psychedelic History Archive, citing the Lilly Papers at Stanford; the catalogue record itself was not reachable. psychedelicarchive.com
• John C. Lilly Papers, 1933–2012, M0786, Department of Special Collections, Stanford University Libraries: the biographical note (the Institute founded in 1959 in Miami and the Virgin Islands), the extent, the acquisition, the folders “Report on Work at St. Thomas July 1963–September 1964,” “Communication Research Institute, audit reports, 1966 December” and “papers to be completed for publication, 1968,” and Series 12.1, the audiotape reels of 1961 to 1968 with their box lists and the note on the digitised reels. oac.cdlib.org, archives.stanford.edu
• Christopher Riley, “The dolphin who loved me: the Nasa-funded project that went wrong,” The Observer, June 8, 2014. Every statement attributed here to Margaret Howe Lovatt in 2014, Mary Lilly’s memory of 1957, Frank Drake, Andy Williamson and Ric O’Barry, “a couple of feet of water,” the six-day week, the autumn of 1966, the Miami bank building, Peter’s death, the Hustler story and the house. In copyright; quoted briefly. theguardian.com
• The Girl Who Talked to Dolphins, produced and directed by Christopher Riley, BBC Scotland for BBC Four, first broadcast June 17, 2014; the Wellcome Collection’s catalogue record and summary. wellcomecollection.org
• John C. Lilly, Lilly on Dolphins: Humans of the Sea, Anchor Press/Doubleday, 1975, introduction: “I no longer wanted to run a concentration camp for my friends” and “five of the dolphins in our eight-dolphin colony committed suicide,” read in the search-inside snippets of the Open Library copy; the sentence “In 1967 to 1968 I decided to stop dolphin work” is quoted at second hand from a reader’s transcription. openlibrary.org, archive.org
• D. Graham Burnett, “A Mind in the Water,” Orion, May–June 2010, and “Off the Dolphin Deep End,” Wilson Quarterly, summer 2010, from his The Sounding of the Whale (Chicago, 2012). The funders, “upward of half a million dollars a year,” the grant examiners at the end of 1965, Nazareth Bay, the Sandoz LSD under an NIMH contract, and the Navy. orionmagazine.org, wilsonquarterly.com
• Bruce Clarke, “John Lilly, The Mind of the Dolphin, and Communication Out of Bounds,” communication +1 3 (1), 2014, open access. “Between 1959 and 1968,” the Institute’s offices and laboratories, and the Career Award “largely financing the CRI since 1962” with its final report. doi.org
• “Anthropologist Gregory Bateson delivers Mandeville lecture,” news release, University of California, San Diego, October 26, 1970: “During 1963-64, Bateson served as Associate Director of the Communication Research Institute on St. Thomas in the Virgin Islands. He joined the staff of the Oceanic Institute in Hawaii in 1965.” library.ucsd.edu
• Ted Nelson, “Script for ‘Man and Dolphin’ film,” January 29, 1963, and “Notes from John Lilly’s Miami Lab, 1963,” for the Institute’s Miami office at Dinner Key and the staff of 1962 to 1963. archive.org, archive.org
• “John C. Lilly,” Wikipedia, for orientation on his later career; the dates used here are those of the Stanford biographical note. en.wikipedia.org