I had been looking into intelligence in nature for eighteen months when a friend called to draw my attention to a recent article in the journal Nature. It claimed that the investigation of plant intelligence is "becoming a serious scientific endeavor" and that scientists are "only now beginning to expose the remarkable complexity of plant behavior." These were the words of Anthony Trewavas, a professor of biology at the University of Edinburgh and a fellow of the Royal Society, the oldest scientific society in Great Britain. According to Trewavas, plants have intentions, make decisions, and compute complex aspects of their environment.
I looked into the research cited by Trewavas and found, to my surprise, that scientists were now saying that plants have senses and can detect a wide variety of external variables, such as light, water, temperature, chemicals, vibrations, gravity, and sounds. They can also react to these factors by changing the way they grow. Plants can forage and compete with one another for resources. When attacked by herbivores, some plants signal for help, releasing chemicals that attract their a.s.sailants" predators. Plants can detect distress signals let off by other plant species and take preventive measures. They can a.s.similate information and respond on the whole-plant level. And they use cell-to-cell communication based on molecular and electrical signals, some of which are remarkably similar to those used by our own neurons. When a plant is damaged, its cells send one another electrical signals just like our own pain messages.
A good part of this knowledge emerged during the 1990s thanks to the development of molecular genetics, which revealed the signals and receptors used by plant cells when they communicate and learn. Anthony Trewavas helped launch this field of investigation with his research on calcium and plant signaling. I contacted him and requested an interview, explaining my purpose. He accepted, and we set up a date.
I arrived in Edinburgh on a cold, stormy January night. As I walked along the streets, I braced myself against the wind and rain. It was my first trip to Scotland. It felt bleak, and I wondered whether I had come to the right place to find out about plant intelligence. I stayed in a hotel on the outskirts of town.
The next morning, the rain had stopped. I made my way over to the university and arrived well ahead of our planned meeting. I wandered around the corridors of the Inst.i.tute of Cell and Molecular Biology, a nothing-special building designed in the 1960s, which now seemed run-down. Corridors in science departments tend to look alike from one country to the next, with drab walls covered with posters announcing conferences or explaining research.
I found Anthony Trewavas in his office on the fourth floor. A tall, balding man, he has piercing light-blue eyes and gray eyebrows. He invited me in and showed me a chair where I could sit down. His office was littered with stacks of journals such as Science and Nature. I glanced at the top file on the nearest pile of doc.u.ments and saw that it was ent.i.tled "Intelligence."
By the time I turned on the tape recorder, Trewavas was already discussing the importance of plant intelligence, saying that scientists have long regarded plants as pa.s.sive creatures, because they lack obvious movement. "Now to my mind, that a.s.sumption is wrong because it requires an equating of movement with intelligence. Movement is an expression of intelligence. It is not intelligence itself. Now, the definitions of intelligence are difficult""
He spoke fluidly, needing no prompting to continue his line of thought. He said he found it necessary to peel away the human aspects that come with the notion of intelligence. In his view, our intelligence did not suddenly appear when we became h.o.m.o sapiens. It evolved from other organisms. Hence the importance of defining intelligence in a way that does not apply exclusively to humans. Trewavas referred to the definition devised in 1974 by New Zealand philosopher and psychologist David Stenhouse, who described intelligence as "adaptively variable behavior within the lifetime of the individual." This can apply to many different organisms and means noninstinctive behavior that maximizes the individual"s fitness.
Trewavas"s desk stood against a bay window overlooking Edinburgh. He sat facing me, with his back to his desk. He looked straight at me as he spoke. His eyes had a piercing quality, but his tone was generous. He said he had spent years pondering the behavior of plants in the light of Stenhouse"s definition. Though most plants do not move at a speed perceptible to the naked eye, they respond as individuals to signals from their environment and develop in adaptively variable ways. Even plants growing in pots inside houses turn their leaves to the light to optimize light collection and send their roots down in the soil and their shoots up into the air. And wild plants manage to compete with other plants for resources. Research now shows that growing shoots can sense neighboring plants. They can detect shifts in infrared light indicative of nearby greenery, predict the consequences of that presence, and take evasive action. Plants can alter the shape and direction of their stems to maintain an optimal position relative to sunlight. They can adjust their growth and development to maximize their fitness in a variable environment. According to Trewavas, this means they are intelligent, if one refers to Stenhouse"s definition.
To ill.u.s.trate his point, Trewavas described the behavior of the stilt palm. This tropical tree has a stem raised on prop roots and moves toward sunlight by growing new prop roots on the sunny side and letting those in the shade die off. By doing this over several months, the stilt palm actually changes places. It "walks" around in this manner, fending off compet.i.tive neighbors and foraging for light, at a speed imperceptible to humans. Trewavas considers this a clear example of "intentional behavior."
Ground ivy is another plant with measurable foraging skills. This perennial weed creeps along the ground as a vine, and when it reaches a patch of optimal size and nutrient content, it puts down roots and generates leaves to catch the light. Scientists recently tested ground ivy in a controlled environment in which nutrients were distributed unevenly. The plant demonstrated that it senses resources by starting to grow roots much earlier in its development in the locations containing nutrients and by skipping over the poorer ground between rich patches. Trewavas finds it "difficult to avoid the conclusion of intention and intelligent choice" in the case of ground ivy.
Such examples cannot be dismissed as preprogrammed rote responses, he said. Rather, they demonstrate plasticity. He explained that an individual plant has an enormous capacity for changing its morphology, its branching structures, to accommodate the environment in which it finds itself. The transformation occurs very slowly from a human point of view, over a period of months, rather than milliseconds. "But the way in which it is conducted and the success with which it has occurred must indicate that a lot of computation goes into the decisions which are actually made, otherwise plants would not dominate this planet in the way that they actually do."
Trewavas had obviously argued in favor of plant intelligence many times. I was willing to consider that Western cultures, and science in particular, had misjudged the vegetal world. But I wondered about the extent of plants" capacities. I asked Trewavas if he thought plants think when they make decisions. He replied that he did not. In his opinion, they compute what is actually going on, then make appropriate responses in terms of what they perceive.
Having answered my question, he continued making the case for plant plasticity. Plants have to gather resources in their local environment while facing compet.i.tion from their neighbors. As they are mainly fixed in one place, the most sensible way any plant can do this is to occupy the s.p.a.ce around itself in an optimal way. A branching structure happens to be the simplest way in which this can be done, and this is the solution plants adopt, both below ground, as they send down roots into the soil to form exploitative tissues, and above ground, as they deploy their leaves to gather the maximum amount of light. To do all this, an individual plant must perceive a gravity vector and align itself correctly. And its actual shape and morphology are determined by the quant.i.ty and quality of light it perceives. For Trewavas, this is "adaptively variable behavior within the lifetime of the individual, i.e., intelligence." Furthermore, individual plants do not choose their environment, as seeds land and germinate where they can. Plants have to grow in a great variety of environments and adjust their structures to optimize their ability to exploit what they find.
Trewavas"s favorite example of vegetal intelligence and plasticity is a parasitic plant called dodder. It moves around by wrapping itself around other plants and correctly estimating their nutritional quality. Within an hour, dodder decides whether to exploit a host or to move on. If it stays, it takes several days before beginning to benefit from its host"s nutrients. But dodder antic.i.p.ates how fruitful its host will be by growing more or less coils. Growing more coils allows greater exploitation; but if the host is poor in nutrients, this wastes precious energy, because dodder lacks leaves and relies on its hosts for water and food. So it has to make correct decisions or face death. Botanist Colleen Kelly, working in the early 1990s, found that dodder correctly a.s.sesses when to eat and when to move on, and that its foraging strategies have the same efficacy as those of animal foragers. And it computes the right choice between close alternatives without the benefit of a brain.
Trewavas described plants as having intention. But I had in mind Jacques Monod"s statement that attributing purpose or goals to nature contradicts the central method of science. According to Monod, studying nature scientifically means ignoring the possibility of intention. I reminded Trewavas of this postulate and added that he seemed to have crossed the line.
He scoffed: "Well, I don"t know how many people actually believe Jacques Monod in that regard. That was an idea that did not really apply to humans, did it? It seemed to devitalize life in my own view. It seemed to indicate that life was solely governed by chance. And animals have foresight. And so do we. And to me, plasticity must be foresight, because it"s the ability to adjust to the particular environmental conditions which you find. If you didn"t have that ability, then you would not be able to accommodate optimally to that. Possessing plasticity is in a sense foresight of the possible conditions in which the plant will actually find itself."
How, then, does a plant make up its mind? I asked. Trewavas replied that he had pondered this question for many years. In 1990, he and his colleagues had a breakthrough. They were studying how plants perceive signals and transmit information internally. Using genetic manipulation, the scientists inserted into tobacco plants a protein that makes them glow when calcium levels rise inside their cells. They suspected changes in cellular calcium concentration to be a major means by which plants perceive external events. To their amazement, they found the tobacco plants responded immediately to touch. Though tobacco is not known to be touch-sensitive, one gentle stroke caused the modified plants to glow with the light produced by the elevation of calcium inside their cells. Trewavas was dazzled by the speed of the response: "It was as fast as we could measure. Whereas I have been telling you that plants only respond in terms of weeks and months, in this case, they were responding in milliseconds to a signal which we knew would later have a morphological effect. If you keep touching a plant, it slows down its growth and it gets thicker."
Trewavas knew that human neurons also use internal calcium elevation when they relay information. Once he saw the speed of the plants" reaction to touch, he started thinking about intelligence. Plants may not have neurons, but their cells use a similar signaling system, he told himself, so they may have the capacity to compute and make decisions.
As I listened to him, I realized that he had firsthand experience of the changes that had swept across contemporary biology in recent decades. He had opened himself to the idea of intelligence in nature. This was a courageous step for a Western scientist. I knew indigenous people in the Amazon who consider it a matter of course that plants have intelligence. But in Western cultures, those who attribute intelligence to plants have long been the objects of ridicule. Until now, scientists, and in particular botanists, had avoided using the words plant intelligence. I wanted to know more about how his thinking had changed and pressed him for details.
Gesturing at the doc.u.ments piled around his office, he said he had read up on a number of different subjects over decades. He described his work method in some detail. "The family used to complain that I would sit in a chair vacantly thinking. I found it very necessary to do. The ideas don"t just come by reading. You have to go away, lie down, sit down, walk about, and let things turn over in your mind. And what I find particularly enjoyable is a problem I"m trying to solve in my own mind. Is there something I can connect together? And I find it"s only by long periods of doing nothing but think that suddenly facts start coming into your mind. And they come together in an interesting combination which enables you to see the possibilities for what plants can actually do." He said the notion of plant intelligence had come to him in this fashion. Intelligence in general was a subject that had interested him for years. So when he saw the connection between plants and calcium, it inevitably led him to think about intelligence.
Trewavas"s intuition about calcium"s role in learning in both animals and plants was confirmed by subsequent research. Scientists recently discovered that when an animal learns to avoid a threat, charged atoms of calcium and specific molecules including enzymes are unleashed inside its neurons. They set about modifying the molecular structure of the channels that span the neurons" outer membranes and control the import and export of charged atoms and molecules. If the threat to the animal persists, its neurons go on to produce proteins that build new connections, or synapses, between neurons. Along with changes in the strength of existing connections, these new synapses give rise to memory, and allow the animal to remember the threat and avoid it.
An a.n.a.logous process occurs in plants. When a plant is threatened, by lack of water, for example, exactly the same atoms and molecules are unleashed inside its cells. And they set off the same reactions, first modifying the same import-export channels, then stimulating the production of proteins if the threat persists. Eventually, the plant modifies its cells and their behavior so that its leaves get smaller, its shoots cease to grow, and its roots extend. These responses minimize further stress and injury to the plant. They also take into account external factors such as nutrients and temperature, as well as the plant"s age and previous history.
Science now indicates that plants, like animals and humans, can learn about the world around them and use cellular mechanisms similar to those we rely on. Plants learn, remember, and decide, without brains.
WE HAD BEEN TALKING for an hour and a half. Trewavas invited me to accompany him to the rooftop cafeteria for a cup of coffee. We wove our way through a labyrinth of corridors and staircases, and through packs of students coming in and out of lectures. The cafeteria was quiet and luminous. It offered a spectacular view of Edinburgh and its surrounding hillsides on a crisp winter day. Trewavas was being generous with his time and knowledge, and was certainly one of the easiest people to interview I had ever met. There had been moments during our conversation when I found it difficult to get a word in edgewise.
Drinking coffee together seemed to be a good time to get more personal. I decided to ask him whether his own behavior toward other species had changed in light of his scientific research. After all, his work showed that we have more in common with plants than most people suspect. He replied that his behavior had not changed much, as he had always respected other species, and had always enjoyed the company of plants and animals. This led him to discuss cruelty toward animals, a much-debated subject in Great Britain. Upon reflection, he realized that his behavior had changed on one count, namely that he had given up fishing. He had come to feel sympathy for the fish, because he could see that a fish on the line is frightened out of its life. Now he considers fishing to be relatively cruel. From his point of view, it is self-evident that animals feel pain. "You throw a fish out of water, and it"s flapping around; well, the reason it"s flapping is because it"s trying to get air. And I suppose I can anthropomorphize that situation and see that I would be doing exactly the same d.a.m.n thing if I was put into water, trying to get air in my lungs, not water. But I like eating fish. I just prefer someone else to catch it. We have to respect the system in which we live, because it will not survive if we don"t respect it. And that"s all there is to it, and I think that is vaguely self-evident. On the other hand, you can"t go overboard about it. We are the important organisms. It"s us discussing the environment and other animals, and not the other way around."
"To our knowledge," I interjected"meaning that we could not be sure that other species were not discussing us. But this did not stop his train of thought. He said that we had to learn to live with other species, and he referred to the work of a fellow member of the Royal Society who had carried out hormonal studies on deer that had been hunted; it showed beyond doubt that these animals were extremely frightened. Trewavas now views hunting animals for pleasure as a lack of respect for life. It was simply untrue, he said, that foxes enjoy a good hunt before being torn to pieces. I found nothing to argue with there.
We returned to his office to wrap up the interview. I asked him about future research on plant intelligence. What remained to be done, he said, was to work out how the whole plant a.s.sesses its circ.u.mstances, makes a decision, and changes what it is doing in response to the environment it perceives. "That requires a lot of communication between the various parts of a plant. It has become an extremely complex area, remarkably complicated. And I can see that we have underestimated this in the past to an enormous extent. People are going to have to keep working on this and try to appreciate that what they are looking at, in fact, is an organism that does exhibit intelligent behavior, and not in ways they normally perceive intelligence."
It was still not clear to me how and where computation occurs in a plant. According to a view Trewavas had expressed in writing, "plant communication is likely to be as complex as within a brain." I told him that when I read that sentence, I pictured the whole plant as a kind of brain.
"Yes, that"s interesting," he said. Then he began comparing the chemical signals used by neurons to those used by plants cells. Some are the same, but others are different. Brain signals tend to be small molecules, whereas plant signals tend to be large and complicated, such as proteins and RNA transcripts. This had only become clear in the last five years, he said. Prior to that, "no one would really believe that proteins would be swimming around a plant providing information." And large molecules can handle large amounts of information, which means there is room for enormous complexity in plant communication. "But you are quite right when you ask about computation: Where does it actually exist? I just don"t know. And the answer is almost certainly: It"s in the whole organism."
Plants do not have brains, so much as act like them.
Later that day, I wandered through the streets of Edinburgh. The clouds had cleared, and the winter sun lay low on the horizon. The city and the volcanic cliffs overlooking it were bathed in pale light. I went over the morning"s conversation with Anthony Trewavas. We humans have different timescales from those in plants. Consequently, we do not see plants move and a.s.sume they are stupid. But this is an incorrect a.s.sumption caused by our animal nature. We do not see them move because we operate in seconds, rather than weeks and months.
I stopped on the sidewalk of the cobblestone street leading up to Edinburgh Castle and remained immobile. I breathed and watched people walk past. I tried shifting to a plant"s timescale, but my thoughts kept racing at animal speed. An image popped into mind of Trewavas sitting in an armchair, not moving, thinking about plants. He was acting like a plant to understand plants, and attributing intelligence to them. Like a shaman, he identified with nature in the name of knowledge. His eyes were shining.
Chapter 8.
SMART SLIME.
Seeing that plants can make decisions led me to look into other cases of intelligent behavior by brainless organisms. I focused on simple species in search of the basic conditions of intelligence.
Amoebas attracted my attention. Their name comes from the Greek amoibe, meaning change. These microscopic single-celled creatures mainly consist of a blob of protoplasm surrounded by a porous, flexible membrane. Amoebas move around by transforming themselves. They change the shape of their bodies by shifting their jellylike contents and stretching their membranes to form extensions known as pseudopods, or "false feet." Amoebas are shape shifters, transformers.
Some amoebas have the capacity to merge with one another to form a single giant cell, with thousands or millions of nuclei. Known as true slime molds, these peculiar unicellular organisms can grow as big as a human hand. And if one of them is diced up, the pieces will put themselves back together. Creeping around slowly and engulfing food along the way, true slime molds act like giant amoebas. There are approximately one thousand species of true slime molds, and they occur around the world, in particular in temperate forests. In their visible, aggregate state, they look like glittering blobs of mucus, or spilled jelly. They can be white, red, orange, or yellow. Typically, a true slime mold changes shape as it crawls over damp wood, leaves, or soil, ingesting bacteria, molds, and fungi. Its entire body is covered by a layer of slime, which it secretes continually and leaves behind as it crawls forward. Though true slime molds are composed of only one large cell, and therefore lack nervous systems and eyes, they can move, navigate, and avoid obstacles. They can also sense food at a distance, and head unerringly toward it.
True slime molds defy categories. They move around to feed themselves, like animals. But they give rise to fruiting bodies containing spores, like fungi. Once their spores disperse to new habitats, they "germinate" into microscopic amoebas. The true slime mold"s life cycle is completed when these tiny amoebas merge into a single, giant cell. True slime molds spend their lives going between two kingdoms, fungi and animal, and between two scales, microscopic and macroscopic.
Scientists recently discovered that true slime mold, Physarum polycephalum, can consistently solve a maze. They found that when separate pieces of this bloblike organism are placed in a maze, they spread out and form a single cell, which fills all the available s.p.a.ce. But when food is placed at the start and end points of the maze, the slime mold withdraws from the dead-end corridors and shrinks its body to a tube spanning the shortest path between food sources. The single-celled slime solves the maze in this way each time it is tested. "This remarkable process of cellular computation implies that cellular materials can show a primitive intelligence," the scientists concluded. The j.a.panese biologist who initiated the experiment, Toshiyuki Nakagaki, declared: "I must recognize that this organism is so clever and cunning." A common view is that intelligence requires a brain. And brains are made of cells. But in this case, a single cell behaves as if it had a brain.
If a single cell of yellowy slime can solve a maze, does this not confirm that the entire edifice of life contains intelligence? I read other publications by Toshiyuki Nakagaki with t.i.tles such as "Amoeboid Organisms May Be More Clever Than We Had Thought" and conclusions such as "I had better change my stupid opinion that a unicellular organism is stupid." I liked what I read so much that I contacted Nakagaki and requested an interview. He replied positively, and I began planning a trip to j.a.pan, a country I had never visited, and where few people speak European languages. I invited along my companion, Beatrice, who has traveled widely in Asia and who is a speech therapist.
In late July, we caught an all-night flight from Switzerland to Tokyo, then flew north to Sapporo, where Nakagaki works as an a.s.sociate professor at Hokkaido University. We arrived in the middle of the afternoon local time, checked into a hotel, had some coffee, then walked around town. The weather was sunny and crisp. Sapporo is modern and easy to get around, with tree-lined avenues. It reminded me of Vancouver. We ended up in a j.a.panese-style Italian restaurant called Africa and drank too much wine.
The following morning, we overslept and barely managed to make our appointment in the hotel lobby. Fortunately, Nakagaki was running late. It was raining outside. He showed up perspiring and carrying an umbrella. He was wearing wire-rimmed gla.s.ses, which suited his oval face. His short black hair was slightly graying on the sides. He seemed to be in his early forties. He dressed in an elegant and relaxed style: a checked shirt, green pants, socks, and thongs. Western clothes, j.a.panese footwear.
We walked under umbrellas as he led us across the campus. There were tall trees and s.p.a.cious lawns between the buildings. Nakagaki explained that an American had founded the University of Hokkaido in the nineteenth century. At one point, he turned to me and said, "Actually, you are not a scientist." I was surprised by his directness. No one had said this to me before; in fact, people often a.s.sume the contrary. But I agreed with him.
We reached the Research Inst.i.tute for Electronic Science, where Nakagaki has his office and laboratory. On entering the building, he asked us to take off our shoes and put on slippers, following j.a.panese custom. As we walked up the stairs to the third floor, he gestured at the walls and said, "This is a cheap building."
Nakagaki"s office appeared bare. It contained a desk, three basic chairs, simple white shelves filled with books, and a writing board. There was a large computer on his desk with a screen showing an e-mail in j.a.panese script. It caught my attention, and I noticed that the keyboard was marked with European characters. I asked how one wrote in j.a.panese on such a computer. He explained that j.a.panese uses three different scripts, including an ideographic script of Chinese origin, an alphabet of syllables to make up for the differences between Chinese and j.a.panese grammars, and a second alphabet of syllables for representing words imported from European languages. He went to the writing board and started showing us the different scripts. Then he returned to the computer and showed how one could shift the keyboard into a mode that allowed one to compose all three j.a.panese scripts. I felt relieved that Nakagaki spoke English.
He asked me to explain my interest in his work. I told him that studying the knowledge of indigenous Amazonians had led me to investigate intelligence in nature. He listened, then commented on the problem Western people have with applying the concept of "intelligence" to nature. He said it was possibly due to the influence of Christianity. I had not turned on the tape recorder yet. I asked him to pause briefly while I did so. Then he resumed and described the conditions in which he and two colleagues"one j.a.panese and one Hungarian"had published their experimental demonstration that a true slime mold can solve a maze. Nakagaki and his j.a.panese colleague did not hesitate to refer to "intelligence" in their conclusion. But the Hungarian co-author proposed to delete the term. The two j.a.panese scientists prevailed, and the journal Nature duly published their paper containing the word intelligence. Much media attention ensued, both in j.a.pan and abroad. Nakagaki said, "I have, in the course of my press interviews about this subject, found myself discussing with foreign reporters just what intelligence is. Whereas j.a.panese reporters were most deeply concerned with the details of just how such an organism was able to solve a maze, those from overseas tended to focus on whether or not the phenomenon represented intelligence."
He attributed this difference to religion. "I got the feeling that some Western people, possibly because of the influence of Christianity, may feel somewhat uncomfortable when faced with the possibility of intelligence other than human." In j.a.pan, he said, people do not hesitate to refer to nature, and even to materials, as intelligent. "In j.a.panese culture, we have a religion of Shinto, which is a sort of animism. So we are likely to accept that everything has spirit, or something like that. This is quite a natural thing for me," he said, laughing.
He got out of his swivel chair, went to the writing board, and marked the j.a.panese term for intelligence: chi-sei, in which chi means to know, to recognize, and sei means property, or character, or feature. Like knowingness, or recognizing-ness. He p.r.o.nounced it CHEE-SAY.
"Chi-sei is the term used to translate the English term intelligence. But I feel there is some difference between these two words, in their background meaning." He wrote the word intelligence on the board: "I feel that behind this term, there is Western Christian culture, in which intelligence is a gift from the G.o.d to humans only." He laughed, then went to his desk and pulled out an article ent.i.tled "Smart Behavior of True Slime Mold in a Labyrinth." He handed it to me, saying it contained his view on the definition of intelligence.
I had already read this article by Nakagaki, in which he reflects on what the true slime mold actually does in the maze. By adjusting its body shape to occupy the shortest route between two food sources, it optimizes its intake of nutrients and its chances of survival. "If the survival mechanism works well even in complicated and difficult situations, then the behavior seems to be smart," Nakagaki writes. "All biological systems must be rather smart. It is not yet known how smart the microorganisms are. In fact, (true slime mold) Physarum"s smartness may be more involved than simply maze solving because life in the wild is more complicated and difficult."
When I first read this article, I wondered what difference Nakagaki made between intelligence and smartness. I put the question to him. "When I use the term smart, Western people agree," he replied, laughing. "Recently I have only used the term smartness."
I asked whether smartness corresponds to the j.a.panese term chi-sei. He said, "Just a moment please" and went back to the drawing board. He seemed at ease standing up, writing out words, and drawing connections between them. He explained that in j.a.pan, people call chemical materials that have functions intelligent materials. But in English, the corresponding term is smart materials. "I didn"t know this correspondence," he said. "I thought Western people used intelligent materials." He a.s.sociated intelligence with "spirit, or mind, or awareness, or something like that," while smartness is "rather neutral, or physical, or well designed." He listed these terms on the board.
I said I understood the term smart to mean flexible and quick when referring to materials.
"Ah, okay, so this word is more appropriate for our study," he said. "Flexibility and adaptability." He wrote both terms under the smartness list.
This prompted me to mention the definition of intelligence used by Anthony Trewavas when referring to plants: "adaptively variable behavior during the lifetime of the individual."
"Yes, yes, yes," he said. "All kinds of organisms have such abilities, adaptability and flexibility. This is true, I believe." He contrasted these abilities to awareness and mind and went on to discuss information processing in biological systems. He wrote the word unconsciousness on the board and said that most information processing in humans occurs at the unconscious level. "So awareness is the small tip of a large mountain. In this sense, all kinds of organisms have a sort of unconscious level of information processing. This ability is very high, higher than we expect."
Nakagaki pulled out a round, plastic dish and handed it to me. It contained the original 3-by-3-centimeter maze in which he and his colleagues had tested the slime mold. It consisted of a negative of the maze cut from a plastic film and superimposed on an agar plate. As true slime molds dislike dry surfaces, they tend to crawl only on the wet, gelatinous agar plate, which the plastic film does not cover.
Then he turned to his computer and showed us some video images of the experiment. First one sees Nakagaki cutting about thirty small pieces from the growing tip of a living slime mold and placing them throughout the maze. As true slime molds move at a speed of about half an inch an hour, it takes a time-lapse camera to reveal their movements. A two-minute sequence concentrating several hours of action shows the bits of slime spreading themselves along the maze"s corridors and blending into one another. They become a single organism, one giant cell covering all available s.p.a.ce within the maze. Nakagaki then places the slime mold"s favorite food, oatmeal, at the start and end points of the maze. Waves start rippling across the yellowish body of the slime mold, emanating from around the oatmeal and splashing down the maze"s corridors. The flat ma.s.s of yellow jelly that makes up the slime mold"s body begins to develop veins that run through the maze. The slime mold ends up withdrawing from blind alleys, avoiding detours, and reducing itself down to a single yellow vein connecting the two food sources by the most direct route.
After seeing these images, I asked Nakagaki if he could show us a living slime mold. He accompanied us out of his office and across the corridor into the storage room for unicellular organisms. The room itself was painted in drab yellow and contained several refrigerators. He opened one and brought out a foot-long plastic container half filled with a bright yellow slime mold. On close inspection, the giant unicellular creature had a solid texture, like mashed potatoes. Nakagaki explained that when a true slime mold lacks water, it goes into a dormant phase during which it becomes dry and can be stored almost indefinitely.
I asked how the idea of putting a true slime mold into a maze first came to him. He said that several years previously, one of his jobs was to feed the laboratory"s slime molds. He usually gave them oat flakes. One day he noticed that if he sprinkled the oat flakes randomly on top of a slime mold, it would form tubes connecting the food sources, and that the tubes were connected to one another in a such a way that the organism derived the maximum amount of nutrients in the minimum amount of time. As Nakagaki has training in mathematics, he began trying "to clarify the smartness of that tube network." He said the point of the maze was to test the expression of that smartness.
We headed back to his office, and he explained that the single-celled slime has the capacity to turn itself into an efficient network of tubes. This is impressive considering that humans have difficulty deducing the shortest connections among just a few locations. He sketched a few examples of tube networks set up by true slime molds. The writing board was starting to look like an evolving road map. He erased old parts and drew over them.
Nakagaki said that a true slime mold turns into an efficient tubing network by contracting and relaxing its body in waves. By varying the rhythm of the contractions, it can move its gelatinous contents either inward or outward. For example, when food is sprinkled on a slime mold, its contractions change drastically. These contraction patterns are self-organized, as there are no leaders or conductors in the protoplasm; rather, parts of the h.o.m.ogeneous slime interact in a synchronized way. Just how this kind of self-organization works is a serious question for mathematics and theoretical physics, according to Nakagaki. "So in this organism, there is no nervous system, no brain, but it has the ability to solve difficult mathematical problems. But the way of computation of this organism is quite unknown," he said.
The rhythmic contractions that ripple across the slime mold and allow it to move are regulated by a complex mechanism that has yet to be elucidated. So far, researchers have determined that different substances partic.i.p.ate in the regulation of these contractions, including charged atoms of calcium, which oscillate. These biochemical oscillators may give rise to waves that propagate through the slime mold"s body and that seem to lead to the development of tubes. But the details remain obscure. Nakagaki thinks the way forward in understanding how a slime mold does what it does is to proceed with mathematical modeling of its behavior, and in particular of its contractions. Understanding what happens in the contraction patterns from a mathematical point of view would allow one to understand how it self-organizes its movements. This, he said, was the main subject of his current research.
I asked how his work had been received by the international scientific community. He said that he goes to international conferences on applied mathematics and physics, and that researchers in these fields have welcomed his work. But he had hardly received any responses from biologists. I found this surprising and asked why he thought it was so. "Recent biologists work on molecular biology," he said. "To such people, it does not matter how the biological system works. They are, in principle, only chemists." He laughed. "But biologists in the field investigating the behavior of animals like my results."
My impression was that an increasing number of scientists were opening up to the idea of intelligence in nature. I asked Nakagaki whether he agreed. He replied that after publishing his research on maze solving by the slime mold, he had become more careful in his use of the term intelligence. Its definition seemed to change from one person to the next, and some critics argued that the slime mold"s behavior could not be considered intelligent because they did not believe it solved the maze by conscious decision.
I asked how those critics could be sure that a slime mold is not conscious.
"I don"t know," he replied. "But, I"ll say it again, consciousness is the small tip of a large mountain." He considered consciousness to be a useful term to refer to self-awareness, as when humans observe themselves observing themselves.
I doubted that introducing concepts of consciousness and self would cast much light on intelligence, if only because the workings of consciousness and the nature of self remain obscure. Nevertheless, Nakagaki"s research showed that the slime mold computes. And many consider computation to be among humanity"s finest intellectual achievements. I asked him about this.
"The slime mold computes," he replied, "but this process corresponds to the unconscious level, I think." He stood up and wrote unconscious level on the board. In his view, most internal information processing takes place on this level, even among human beings. "I doubt anyone could explain how it is their body maintains balance when they ride a bicycle. While we are riding, our body just naturally performs the calculations required to solve the equation. It would be quite difficult for us to clearly define these on the conscious level, and were one able to do so and publish the method employed, it would undoubtedly be an important contribution to the scientific literature." For Nakagaki, all living organisms have unconscious information-processing mechanisms. Whether this const.i.tutes intelligence is a matter of debate. His research aims at clarifying these mechanisms, he said, if possible at a material level, in order to find out whether or not single-celled creatures possess intelligence. In this effort, he considers the slime mold to be an ideal subject.
Having spent the afternoon talking, we went out to dinner. Nakagaki invited along his wife, Yuka, and their three-year-old son, Gen-ichiro. We went to a restaurant specializing in traditional j.a.panese cooking and sat together around a low table in a room part.i.tioned off from others by bamboo walls. Yuka had worked as a travel agent for ten years. She spoke with enthusiasm, and in fluent English, about South Korea, one of her favorite countries to visit. Gen-ichiro played quietly with his mother"s cell phone. Though we drank a number of gla.s.ses of sake, I still had some questions. In particular, I wanted to know what Nakagaki thought about the importance of studying intelligence in nature. He replied that it is "one of the most important questions in science."
I agreed but said that, until recently, most scientists had held the opinion that nature lacks intelligence.
"So, this opinion is wrong. This is obvious," he said. "Most scientists are surely ill informed on this question. They only think about their own subject. Apart from their own subject, they are ill informed."
He looked straight at me from across the table and added, "You think about intelligence in nature, and you investigate many cases of research describing intelligence in nature. So you know more about this intelligence than I do. So you are the specialist on the problem of intelligence in nature. Whether you are a scientist or not does not matter. Since the times of Greek philosophy, we have basic questions on the mind and intelligence. Archimedes and Pythagoras thought about these serious problems. Descartes also thought about them. In this time, only a few people think about this serious problem. We do not have to share the opinions of most scientists."
After the discussion with Nakagaki, I thought about the concept of chi-sei. He said that j.a.panese people did not question applying this term to the maze-solving slime mold. This was perhaps a concept I needed. Intelligence had been defined in too many different ways and had become a loaded word. And smartness commonly means cleanliness, tidiness, and elegance, which weakened its pertinence to my investigation. When a true slime mold solves a maze, it demonstrates a capacity to recognize its situation, to know. And if a true slime mold has chi-sei, what living ent.i.ty does not?
Chapter 9.
j.a.pANESE b.u.t.tERFLY MACHINES.
After hiking up a smoking volcano near Sapporo, Beatrice and I headed south to Kyoto, the historical center of j.a.panese culture. Kyoto is hot and muggy in the summer. It also has two thousand temples. We spent several days seeing the sights. We walked along the Path of Philosophy, which follows a ca.n.a.l lined with cherry trees. We visited the Golden Temple Kinkaku-ji in the rain. We strolled through manicured gardens with moss carpets and ponds filled with sacred carp. One sign with an English translation posted at the entrance of a temple explained that Zen gardens are "compressed nature." Another sign above a small exhibit of moss samples stated: "Very Important Moss (like VIP)." Paying attention to details in nature appeared to be a j.a.panese talent.
We caught a train from Kyoto to Tokyo and settled into a small hotel downtown. The sheer size of Tokyo takes getting used to. No sooner had we found our bearings than the first typhoon of the season blew in. Dark clouds filled the sky, and gales of wind blasted down the avenues. Almost horizontal sheets of rain poured down. People in the streets braced themselves and walked with their umbrellas directed against the winds.
The next day, the typhoon was still raging, and we traveled to Yokohama, the country"s second biggest city, which now forms an uninterrupted megalopolis with Tokyo. I had an appointment at the University of Yokohama City with Kentaro Arikawa, a professor who has been studying b.u.t.terfly neurology for twenty-five years. Arikawa is the scientist who discovered that b.u.t.terflies have color vision, and that their tiny brains contain sophisticated visual systems. He also discovered that b.u.t.terflies have eyes on their genitals.
The Tokyo subway system is mainly signposted in j.a.panese, and labyrinthine. We ended up finding the over-ground line to Yokohama, which we rode for an hour through an unending urban landscape. The train shook from the storm raging all around us. Once we reached our final destination, I called Kentaro Arikawa from a public phone outside the station, as he had instructed me to. A few minutes later, he appeared driving a gray car and flashed his lights in our direction. We were easy to recognize as the only gaijin, or foreigners, in the vicinity. We rushed through the downpour and got into his car as quickly as possible. We shook hands, then Arikawa drove off saying that we did not have far to go.
I sat in the front seat and wiped the rain from my face. Arikawa was a lanky man with short black hair, wire-rimmed gla.s.ses, and a kind, gentle face. He was in his mid-forties. He wore a a short-sleeved shirt, dark pants, leather shoes, and a big watch that looked suited to underwater diving. After a short drive we reached the campus of Yokohama City University and pulled up in front of the Graduate School of Integrated Science, where Arikawa teaches and conducts research. As we dashed from the car to the main entrance, I asked him what b.u.t.terflies do during typhoons. "They hide in holes in trees," he said, "or under leaves."
This time we did not take off our shoes. We walked over to the elevator, went up to the fifth floor, and proceeded into Arikawa"s office. He invited us to sit around a comfortable table and offered to make some tea. I explained my interest in his work by describing my investigation and saying I saw clear indications of intelligence on nearly all levels of nature, including in plants.
"I don"t know much about plants," he said, "but our intelligence must have originated from animals which were our ancestors. So intelligence, the mechanism of making decisions, must exist in present-day animals. And as you say, it is widespread, even in b.u.t.terflies." He described the work he and his colleagues are conducting, looking at the capacity of b.u.t.terflies to see colors: "We have already found an enormous complexity in the eye. And of course we are looking at conscious behavior, and we showed that they can clearly see colors and have color constancy."
Arikawa explained color constancy by giving the example of a human observer who sees a red apple as red in both sunshine and regular room light, though the spectral contents of sun and room light are very different; in such a case, the subjective experience of red remains the same, because the observer"s brain adjusts its perception of the wavelengths reaching the eyes. This is color constancy. It turns out that the microbrains of b.u.t.terflies are also capable of this feat.
Arikawa pulled out a black page showing a series of colored patches and began explaining how he and several colleagues had demonstrated that j.a.panese yellow swallowtail b.u.t.terflies have color vision and color constancy. The scientists trained the b.u.t.terflies to feed on sugared water placed on a patch of a particular color in a cage set in the laboratory. Then they presented the b.u.t.terflies with the training color randomly positioned within an array of patches and devoid of sugared water. The b.u.t.terflies selected the training color reliably among different colors, including a variety of shades of gray. They also selected it under different-colored lights, showing color constancy. b.u.t.terflies must be able to see colors in order to recognize suitable flowers for feeding in the field. They use color information to collect food. And because food must be food, under direct sunshine or in the woods or anywhere else, color constancy is important to b.u.t.terflies.
Arikawa and his colleagues also demonstrated in the course of their studies that the retina of the swallowtail b.u.t.terfly has at least five different types of spectral receptor: ultraviolet, violet, blue, green, and red. They recently found a sixth receptor, which is broadband, and probably works as a general luminosity detector. In comparison, humans have only three types of spectral receptors: red, green, and blue. Arikawa and his colleagues concluded: "The extremely richly endowed visual system of b.u.t.terflies evidently provides these animals with a versatile information-processing apparatus."
Astonishingly, the tiny brain of a b.u.t.terfly is equipped with a system of color vision that is superior in some respects to our own.
Ultraviolet photoreceptors serve several purposes. They enable b.u.t.terflies to see flowers that have pigmented ultraviolet spots indicative of nectar and pollen within. They also allow male b.u.t.terflies to detect the distinctive ultraviolet stripes on the hind wings of female b.u.t.terflies, which facilitates courtship and mating. Sometimes nature uses signs that human eyes cannot detect.
b.u.t.terfly visual systems develop during metamorphosis, when young b.u.t.terflies are still full-grown caterpillars undergoing self-transformation in the pupa. While caterpillars have six simple eyes on each side of the head, b.u.t.terflies develop an additional pair of large, compound eyes. The simple eyes of caterpillars have only three kinds of photoreceptors, while the compound eyes of b.u.t.terflies have twice as many. b.u.t.terflies are transformers. They do not sprout just wings in the pupa but brand-new eyes as well.