The Pendulum of Educational Progress

By Thomas R. Hoerr

Progress is rarely smooth on important and complex issues. Steps forward are followed by steps backwards, and we can hope (we must hope!) that the forward steps are longer than those in retreat. This is certainly the case when considering an incredibly complex and very relevant issue, the definition of intelligence. Despite resistance to change, new understandings about child development, growth, and intellect must push against decades of traditional thinking and comfortable practices.

New City School, St Louis, Missouri – an MI school

We saw this tension in the 1980s when Frames Of Mind: the theory of Multiple Intelligences was first published. Multiple intelligences (MI) redefined–expanded–the traditional view of intelligence as a single entity that could be determined by a score on a test. Although Howard Gardner wrote about MI with psychologists in mind, it was elementary and secondary educators who embraced this new way of understanding students’ abilities and potential. Instead of assessing students to determine a hierarchy of intelligence by asking, “Who’s the smartest?” the question became “How is each child smart?” This expanded definition of intelligence valued problem solving skills and abilities in art, music, athletics, nature, working with others, and knowing oneself. A pragmatic approach to intelligence, MI captured the range of abilities that can lead to success in life.

Visiting Hamilton High School in Hamilton, Ohio, Jan. 8, 2002, President George W. Bush signs into law the "No Child Left Behind Act." (Wikimedia)

The impact of MI in a school was powerful. I know this from working in an MI school. Recognizing MI, educators began to reconsider curriculum and pedagogy in ways which enabled students to use all of their intelligences in solving problems; students learned and they learned with joy.

But then the U.S. Congress passed No Child Left Behind (NCLB) legislation in 2001. This meant that student success on standardized tests–which focused only on the linguistic and logical-mathematical intelligences–became the criterion for measuring achievement; failure to excel on standardized measures threatened jobs and the viability of schools. This narrow pathway to success ignored the intelligences and passions of many students; it told them that they weren’t smart. There was–and is–much pushback by K-12 educators about the narrowness of standardized tests as the criterion (the criterion, not one of many criteria), but test scores remained the altar.

In an attempt to pursue a wider understanding of student support, the Every Student Succeeds Act (ESSA) became law in 2015. Schools were given more flexibility in measuring student progress, a clear push-back against NCLB. At the same time, ASCD popularized the term “whole child” to remind everyone that students were more than their test scores. Today, all 50 states have SEL (social and emotional learning) standards, and CASEL (the Consortium for Academic, Social, and Emotional Learning is a major force in looking at student capacities and potential. These are significant steps forward in recognizing that problem-solving, the definition of intelligence, is not limited to reading, writing, and calculating.

But recall my opening comment about progress not being smooth. Today, there is resistance to viewing student potential and problem-solving in a broader context. Teachers and administrators share that rhetoric aside, what counts, what they are asked about by supervisors and, often parents, are their students’ standardized test scores. This narrow focus is reinforced by today’s political milieu. Fatuous claims are being made about the intelligence of Haitians, for example, in arguments to limit immigration. Of course, these same arguments about the lower intelligence of ethnic groups were used about Irish and Italian immigrants as well as African Americans in the past.

The notion of the bell-shaped curve, the idea that virtually all human variations fall neatly along a continuum, with the bulk of a characteristic in the middle and an equal proportion falling along each side, is false. Charles Murray and James Herrnstein proclaimed this in their book, The Bell Curve: Intelligence and Class Structure in American Life, citing the model as evidence for the intellectual superiority of some groups. But in More Like Us, James Fallows noted that the bell-shape curve that exists in standard intelligence testing occurs because the tests are designed to elicit that result! Test items which are chosen to lead to that bell-shaped curve, thus reinforcing the model which many people accept despite its limitations.

While all states now note SEL as a positive, an editor recently told me that sales of a book for teachers were limited because the author used the term “SEL” in the text. As another example of our internecine politics, references to DEI (diversity, equity, and inclusion) issues are banned from curriculum in some states. Rulings by the U.S. Supreme Court have made it more difficult for universities to consider all of a child’s background and potential in determining enrollment. Yet some politicians and school leaders hide behind the purported objectivity of standardized scores, ignoring the strengths that students have which are not evident in a standardized test.

Indeed, the positive correlation among socio-economic status, race, and student achievement affirms a narrow definition of intelligence and the socio-economic and racial hierarchy within our country. This revanchist view ignores the potential of too many students. Steps forward are followed by steps backwards and, in turn, those steps backwards are followed by progress. I remain confident schools will recognize and address a wider range of student capacities, multiple intelligences; indeed, AI will increase the speed at which that happens. I just want the pendulum to swing to the benefit of students as soon as possible.

Who Owns Intelligence? Reflections After a Quarter Century

© October 23, 2024 Howard Gardner, Shinri Furuzawa, Annie Stachura

Introduction

At the very end of the 20th century, The Atlantic Monthly (as it was then named) invited Howard to submit an essay about the psychological concept of intelligence. Under the provocative title “Who Owns Intelligence?,” Howard contended that for most of the 20th century, psychologists—and chief among them, psychometricians—essentially owned the concept of intelligence. On their formulation, intelligence was a singular capacity, presumably largely inherited, along which human beings could be reliably arrayed. An IQ of 100 denoted the mean/average for members of our species; those scoring above 120 or 130 (hence “smart”) were likely to become leaders and significant contributors to their respective societies; in contrast, those at 80 or below (hence “dumb”) were deemed poorly equipped to thrive and typically in need of supportive interventions. It’s important to point out that many psychologists—and most psychometricians—still cling to this view. (For a critique, see here.)

Even a quarter of a century ago, this scholarly consensus was being increasingly questioned or challenged. With his theory of multiple intelligences—postulating at least seven distinct intellectual faculties—Howard was perhaps the chief challenger; but among the critics were earlier psychologists like Louis Thurstone (seven “vectors of intelligence”) and J.P. Guilford (up to 150 “mental abilities”) as well as contemporaries like Robert Sternberg (whose triarchic theory posits analytic, practical, and creative forms of intelligence).

The hegemony of the standard-IQ saga was clearly broken when psychologist-turned-science journalist Daniel Goleman detailed the nature and significance of emotional intelligence (soon dubbed by Newsweek magazine as “EQ”). Not only did Goleman outline psychometric approaches to the ascertainment of EQ; but he delineated the importance of EQ for work, social life, leadership, collegiality, friendship—indeed overall thriving. There are now scores (if not thousands) of books, programs, and tests all burnishing the significance, accessibility, and enhancing of this form of intelligence. 

Daniel Goleman

EQ exploded the fortress that had hitherto blocked the positing of additional forms of human intelligence. Nearly every week, Howard receives correspondence from individuals seeking to posit or nominate additional forms of intelligence—humor, mechanical, sexual, computer, culinary—you think of it, someone names it, someone purports to measure it, the “intelligences game” goes on. In response, Howard points out that it’s easy to come up with candidate intelligences; but his original list of seven (multiple) intelligences was restricted to those candidates that fulfill eight specified criteria ranging from a neurological basis to cross-cultural visibility and viability. (And indeed, in over forty years, Howard has added only one intelligence—that of the naturalist—to the original septet). By the same token, those who would posit additional intelligences need to show how these candidate faculties could be ascertained and whether indeed they appear to be distinct from those already proposed by the aforementioned scholars. Coming up with candidate intelligences is a parlor game; convincing a skeptic (or a scholar) of their plausibility requires serious work—and the willingness to reach a negative conclusion. 

The previous paragraphs could have been drafted a quarter of a century ago. But in the intervening period, there have been serious efforts to expand significantly—perhaps to the breaking point—the concept of intelligence. (For an example of an effort that goes too far, see Sparavigna, 2014. This authority extends the descriptor and the location of “intelligence” to include the Earth, various planetary systems, and a bridge between heaven and Earth.)

In what follows, we undertake four tasks: 

I. Reviewing the claims and evidence for forms of intelligence in non-human animals

II. Reviewing the claims and evidence for forms of intelligence in plants

III. Reviewing the claims and evidence for forms of computational intelligence—in particular, those displayed by ChatGPT and other Large Language Models (LLM)

IV. Stepping back, we propose a novel way of conceptualizing intelligence—one that acknowledges these disparate claims—while conceding that the intelligence(s) of human beings may still merit special status. Specifically, building on an approach developed by philosopher Nelson Goodman, we propose several “symptoms of intelligence;” we suggest that proposed candidates be judged on the extent to which they exhibit—or fail to exhibit—these symptoms.

I. Reviewing the Claims and Evidence for Forms of Intelligence in Non-Human Animals

Research on animal intelligence, or the field of animal cognition, has increased dramatically in recent years. Almost every day, news articles report discoveries that seem to indicate a previously unknown intelligence in a non-human species with implications for our understanding of animal cognition and consciousness.

Historically, however, in Western philosophy non-human animals have not been considered intelligent, nor even conscious and capable of feeling. Descartes famously believed that animals were like “automata,” incapable of feeling pain. This view was challenged by philosophers of the 18th century Enlightenment such as Locke, Hume, and Rousseau. These influential thinkers contended that animals were similar enough to humans that, though they might not be rational, they likely had similar perceptions and understanding based on instinct. A century later, Charles Darwin discerned evolutionary continuity and similarities between humans and animals in intelligence and consciousness.

A swing against support for animal intelligence came in the early to mid-20th century when behaviorists, such as B. F. Skinner, argued that all animal behaviors were simply due to varying schedules of operant conditioning. On this analysis, animals were simply responding to stimuli from the environment without any conscious decision making or what might appear to be intellectual engagements. Some even extended (and still extend) this way of thinking to human beings. It’s worth noting that behaviorist views were based chiefly on observations of solitary animals performing in the lab, far from the animal’s natural environments and normal living conditions. (If you make a leap to the single human subject taking an IQ test in a monitored auditorium, we won’t protest!)

Consciousness and self-awareness

Behaviorist ideas were challenged when the fields of psychology and neuroscience came together in the (the so-called) cognitive sciences, and various fields of biology and genetics merged with the Modern Evolutionary Synthesis. A clearer understanding of the mind and mental processes, together with a more integrated understanding of evolution, has led to a more open-minded approach to the possibilities of animal cognition and intelligence.

In more recent decades, scientists have begun to consider non-human animals as having considerable cognitive capacities—and even consciousness and self-consciousness. Indeed, in 2012, a group of neuroscientists signed the Cambridge Declaration of Consciousness. This declaration “unequivocally” stated that “humans are not unique in possessing the neurological substrates that generate consciousness. Non-human animals, including all mammals and birds, and many other creatures, including octopuses, also possess these neural substrates.”

As of today, many species are believed to have fundamental self-awareness, knowing that they are distinct beings and not the same entities as their surroundings or their conspecifics. The mirror test provides one form of evidence. This test assesses whether an animal investigates and attempts to remove or touch a mark made by researchers on the animal’s body that is only visible in a mirror. Should an animal demonstrate awareness that the reflection is of itself, this behavior signals a level of self-awareness or at least self-recognition—and, possibly, of other valued cognitive processes. Animals that have passed the mirror test include primates (such as bonobos, chimpanzees, orangutans), dolphins, birds (such as magpies, parrots), elephants, and cleaner wrasse fish. Other possible indications of self-awareness include the ability of boa constrictors to choose the correct size hole their bodies will fit through before and after eating; or dogs recognizing their own scent as distinguishable from the identifiable scent of other dogs.

Neuroscientist Antonio Damasio argues that all living organisms may be able to sense their environment and have covert or unminded intelligence. Bacteria and unicellular organisms are capable of sensing and responding to stimuli such as light or heat and this does not require consciousness or knowing how and why they problem solve. He credits more complex organisms with higher levels of consciousness and explicit and overt intelligence, enabled by a nervous system and neurons. This form of intelligence requires feelings and the forming of imagistic patterns in the mind. On this account, consciousness in animals exists on a spectrum, with some species having more complex and reflective metacognitive capacities than others.

The importance of neurons, if not a large brain

With advances in neuroimaging and genetic analysis, we are now able to relate putative differences in intellectual capacity to different brain structures. The sheer size of an animal’s brain was once considered the indication of its intelligence, with humans having of course the largest brain relative to body size. In the 1970s, however, ethologists and neuroscientists determined that dolphins, crows, and certain primates, display comparably high levels of problem solving and social intelligence despite having distinctly differently-sized brains. Neuron density, brain organization, and/or connectivity patterns may be more important than sheer size.

Indeed, neurons do not even have to be located within the brain. We now know that octopuses have roughly two thirds of their neurons outside the central brain and distributed throughout the body, especially in the arms. In fact, the number of neurons in each octopus arm exceeds the number in the entire brain of comparably complex animals.

Some researchers have argued that super organisms (a group of synergistically-interacting organisms of the same species), such as slime mold or ants, could be described as having a “liquid brain.” Single-celled organisms without a central nervous system or brain, slime molds can collectively solve problems, learn from experiences, make decisions, process information and adapt to their environment in highly efficient ways. Similarly, army ants have small individual brains with few neurons, but they collectively display intelligence and are capable of solving mazes and mapping networks. They efficiently coordinate and modify their behavior using tactical cues, such as drumming on each other’s heads, and pheromones to transmit information. “Ant colony optimization” can be mathematically modeled to solve complex problems such as connecting two points in a network by the shortest route. Borrowing the analytic approach of Damasio, non-neuronal superorganisms such as slime molds, may be considered to have unminded intelligence; but they do not possess minds or consciousness and are only displaying problem-solving skills that allow efficient existence.

Problem solving

When it comes to using our explicit intellectual endowment to solve problems, one aspect of human exceptionality was long assumed to be our use of tools. Since the 1960s, however, we have recognized that many other species both create and use tools to solve problems. Jane Goodall famously observed chimpanzees using sticks to “fish,” stripping the leaves and sticking them into termite mounds to extract and eat the termites. Other examples include bottlenose dolphins using marine sponges to protect their rostrums (beaks) while foraging the seabed and using empty shells to trap fish. Octopuses carry coconut shells as portable shelters to protect from predators and stack rocks as barriers for additional protection. Using and creating tools to find innovative solutions to problems that are of value to the species may even be considered a form of animal creativity.

Some species also exhibit problem-solving skills that involve using both prior learned knowledge and future inference. For example, crows place nuts on roads and wait for cars to drive over them. In urban environments, racoons learn to open locks, garbage cans, garage doors, and windows, persistently working on a problem for hours and returning to it.

The problem-solving skills of various species have also been demonstrated in the lab. The ability of rats to solve mazes is well known. Octopuses can also navigate mazes by learning and remembering routes. Exhibiting long-term memory, they recall solutions to puzzles when faced subsequently with similar challenges—they can even open childproof jars containing food and have been known to escape their tanks. Through observation and imitation of conspecifics, octopuses display social learning. Corvids can solve problems and puzzles with up to 8 steps.

Some animals solve problems using arithmetic and estimation of quantity. Bees, for example, count landmarks to mark distance to navigate routes to food sources (note: many known human languages don’t count above four). Guppy fish and angelfish distinguish between group sizes, choosing larger groups for better chances of survival. In the lab, chimpanzees have demonstrated better short-term memory than humans for numbers. Numerical competence demonstrates once again that certain proposed intelligences are not unique to humans.

Cultural transmission

Animals also display a form of intelligence through their ability to share and teach acquired knowledge across generations, a form of cognition also previously thought to be unique to humans. Bottlenose dolphin mothers, for example, teach their young how to use sponges when foraging, a skill primarily passed from mother to daughter. Matriarchs in elephant herds take charge of guiding and teaching young elephants migration routes, as well as navigation and survival skills. Chimpanzees have been observed teaching their offspring which specific plants have medicinal properties and how to use them. Mothers guide their young to foraging sites and demonstrate how to use the plants—for example by chewing first or swallowing whole to help expel intestinal parasites. Crows share knowledge gained about specific “threatening humans,” which enhances the survival of the group.

Communication

When it comes to communication (such as that exhibited in this article!), as far as we are aware, other animals do not exhibit nuanced or abstract forms of thought. We don’t know whether animals communicate about past or future events. Though they may transmit messages on a simple level, their communication appears to lack the grammatical complexity or depth of human language.

And yet, in some ways, animal communication resembles human language. Just as humans have distinct societies with different languages, orca groups have different “dialects;” variance across groups suggests distinct “group cultures.” Some bird species—like songbirds and parrots—exhibit syntactic intention as they arrange sounds to convey different meanings as in human language. Japanese tits, for example, can use an “ABC” call sequence as a predator alert, which when adding “D” may indicate the type of threat. Primates also have a rich vocal repertoire used to communicate, for example, about food, danger, or play. The neural mechanisms that enable vocal production are similar to those found in humans and may illuminate the evolutionary origins of human speech.

Scientists are only beginning to discover sophisticated communication within certain species—it may be far more complex than observers had once thought. It appears that some species use forms of communication—such as vibrations, chemical signals, or sign language—that prove difficult for humans to discern or perceive. For example, dolphins do not just “hear” sound; they “feel” sound through their tissue and surrounding water. This richer experience—both tactile and acoustic—may well convey nuanced information. Sperm whales use click vocalization to communicate complex ideas—such as information on the location of feeding grounds, migration routes, or approaching dangers—and perhaps even to share emotions. They use nasal passages and a large fatty organ in their foreheads (the melon) to amplify and direct large sounds over long distances. The Cetacean Translation Initiative is currently using AI to decode sperm whale clicks—much more complex than previously realized. The speed and timing of clicks correspond to human vowel duration and pitch, and the timbre and harmonics of clicks correspond to human resonant frequencies during singing and talking.

Elephants also have a highly rich communication system. We now know this involves not only vocalizations, body language, touch, chemical cues through pheromones and gland secretions, seismic signals produced by foot stamping and other movements—but also infrasound. Low frequency infrasound waves produced in the larynx and modulated through the vocal tract and skull can travel several miles through the air and even greater distances underground. Infrasound is sensed through specialized cells in the feet as well as through the ears and other parts of the body.

Some animal species have been taught to communicate successfully with humans as demonstrated by Koko the gorilla trained by Francine Patterson. Koko learned over 1000 signs in American Sign Language (ASL) and could be said to “understand” English. Washoe the chimpanzee, taught by Allen and Beatrix Gardner, also used ASL and could form simple sentences. Irene Pepperberg trained an African gray parrot named Alex who could name over 100 objects and use phrases and sentences. Alex demonstrated advanced cognitive skills thought to exist previously in only primates and humans. Other species taught to communicate with humans include crows and ravens, dolphins, elephants, sea lions, and dogs.

Complex social behavior

Researchers are not only learning to decode the communication systems of diverse species, but they are also beginning to uncover social behaviors that are more complex than previously assumed. Skeptics of animal cognition have argued that attributing complex social behaviors to animals is anthropomorphizing—projecting human motivations onto other species and overinterpreting their behavior.

© Nachiketha Sharma for The New York Times

We now know, however, that chimpanzees are capable of perspective taking, or understanding the goals and intentions of others—distinguishing whether another chimpanzee is unwilling or unable to cooperate. Scrub jays practice deception—they “know” when another scrub jay is trying to steal their cache of food and will “pretend” to hide it in one place while actually hiding it elsewhere; or they “feign” indifference when discovering food so as not to attract attention and return to it later when the coast is clear. Elephants display empathy, another form of perspective taking; they offer comfort to other elephants who are grieving or assist others in difficulty. Such behaviors require the ability to recognize mental states in others and understand their intentions.

Perception

One of the challenges in understanding animal cognition is that many species perceive the world in ways fundamentally different from those exhibited by humans. Some have electroreception, such as glass knifefish that use distortions in electric fields to navigate and hunt. Honeybees see a much wider UV spectrum perceiving ultraviolet light and snakes can perceive infrared radiation. Some birds, fish, and reptiles have trichromatic vision—they can see additional color channels beyond the red, green and blue perceivable by humans. Other animals, such as bats and dolphins, can perceive high-frequency ultrasound, while some, such as elephants and whales, perceive low-frequency infrasound. Bats use echolocation calls beyond the range of human auditory perception. Integrated visual and auditory systems process returning calls to map the environment in considerable detail, allowing bats to navigate and hunt in the dark.

Migratory species, such as Arctic terns and monarch butterflies, are able to sense magnetic fields and use cues from the sun, stars, and polarized light to navigate many thousands of miles. Sea turtles are able to detect the distinct magnetic signatures of different beaches; they use such cues to cross oceans navigating back to beaches where they were born to lay eggs, possibly also aided by olfactory cues.

Dogs may well deserve their own paragraph! They have an ability to use olfactory senses to map a rich awareness of their environments, not just of the present but through lingering scent trails—of the past. Dogs also exhibit both inter-species and intra-species differences, with variations in performance according to factors such as breed, training, motivation, health, and temperament. As an example: bloodhounds have an extremely large array of olfactory receptors (possessing around 300 million) which enhances their olfactory abilities (humans possess a mere 5-6 million)—along with a very good working memory. In contrast, companion dogs have poorer working memory but are adept at interpreting social signals from their owners.

The diverse perceptions of the world by different species demonstrate the complexity of animal cognition and range of adaptations that allow thriving in different environments. Other animal species experience the world in different ways to humans and process information in ways we are unable to appreciate without technological assistance.

Reflections on the Posited “Multiple Intelligences” in Other Species

We have discussed some of the advances in our understanding of animal cognition. We now know that intelligence is not determined by brain size or configuration; that many animals exhibit a level of consciousness and self-awareness, along with considerable problem-solving skills and creative tool use. Animals have means of communication and ways of perceiving the world that humans have not yet decoded or cannot decipher without the help of technology. They also have complex social behavior and ways of interacting that seem both familiar and alien to us as humans.

In speaking of the familiar, an anthropocentric focus can lead to biases, but also provides a useful lens for understanding other species. Howard developed the theory of multiple intelligences to describe the range of human capacities; here we reflect on the existence and purview of those intelligences in other animal species.

Linguistic intelligence: Many species have complex and varied forms of communication, though these may be very different from human language.

Logical-mathematical intelligence: Non-human species do not use numerical symbols or understand higher mathematical concepts; however, many do exhibit some numerical competence.

Interpersonal intelligence: Some species have “Theory of Mind” and exhibit norms of social behavior. They have expectations of fair treatment enforced by punishments and rewards. They also practice altruism (with an expectation of reciprocity) and empathy or compassion—in this respect, they may be more exemplary than some human beings! Prosocial groups tend to be more successful as evolution rewards such behavior.

Intrapersonal intelligence: Many species display a degree of self-awareness—though we would certainly not claim that they would seek—or benefit from—any form of psychotherapy!

Musical intelligence: Animals do not compose symphonies or learn to play instruments for their enjoyment. They do, however, produce sounds which sound similar to music to the human ear. Nightingales and canaries produce melodic and musical songs, and the songs of humpback whales have complex themes, phrases, and subphrases which can continue over several hours. The purpose of these songs is presumably communication—as examples, to attract mates or ward off predators. Even though humans may classify these sounds as musical, the ability to produce and decode these sounds may in fact be closer to linguistic intelligence.

Bodily-kinesthetic intelligence: Specialized body movements provide an evolutionary advantage to a species. The various physical skills and coordination required for different species to survive in their environments can be deemed a form of intelligence. Bodily-kinesthetic intelligence also allows the use of tools for problem solving. Many species are capable of being trained to perform complex physical feats, such as the synchronized acrobatic tricks of dolphins, or dressage and show-jumping events with horses.

Spatial intelligence: We do not know whether animals have the capacity to visualize and manipulate objects and spatial dimensions in ways similar to those exhibited by humans. But many animal species demonstrate high levels of spatial intelligence—enabling them to locate food sources, avoid predators, find mates, and navigate over long distances. Clark’s nutcracker birds can remember 2000 cached food sites for almost a year and have well-developed hippocampus regions associated with spatial navigation and cognitive mapping.

Naturalist intelligence: As non-human animals interact with the environment for survival, they have the intelligence to recognize, categorize, and distinguish other species and various aspects of their ecological niches. Many animal species have the ability to know their environment through different forms of perception—not simply by sight, but by smell, or magnetic field. For survival, plant species might be distinguished for their different nutritional value, and different species of both prey and predators are recognized. Indeed, many animal species may well surpass the naturalist intelligence of human beings who—long living in urban environments—have lost the ability to recognize a wide variety of flora and fauna….one reason for providing summer camping experiences to urban youth!

We must be wary of human exceptionalism; there is much we do not yet know regarding animal intelligence. Primatologist Frans de Waal challenged us with his book entitled: Are We Smart Enough to Know How Smart Animals Are? Human intelligence has led us to develop moral systems, religion, art, science, and philosophy. However, we also use our intelligences to rationalize war and genocide and cause irreparable damage to the planet.

Intelligence is a valorized term—therefore, an acknowledgement of intelligence in non-human animals should lead us to value diverse species and appreciate a range of intelligences. There is also an ethical implication in recognizing that non-human species exhibit varieties of intelligence and consciousness. It’s time to reconsider our relationship with other animals and protect their interests on the planet we share. Many countries have enacted animal protection and welfare laws, and organizations such as the Nonhuman Rights Project advocate for the extension of legal personhood to certain animals such as primates and elephants. Such efforts are, however, controversial; reflection on animal rights continues to evolve as we advance our understanding of animal cognition and consciousness.

II. Reviewing the Claims and Evidence for Forms of Intelligence in Plants

In 2019, Richard Powers’s novel, The Overstory, dubbed by The Washington Post as “a sprawling epic about the wonderful life and alarming death of trees,” won the Pulitzer Prize for Fiction. The story follows nine Americans whose lives are touched by trees—these characters ultimately come together to address the destructive impacts of deforestation. The trees in Powers’ epic novel are often personified. They laugh, remember, take care, know, offer, and speak—they are depicted as both memory and memory-keeper, both tethered and untethered to our human stories.

Imbuing the world of plants with human qualities is common practice in literature. But to what extent can we properly call plants and plant behavior intelligent in the real world? Particularly in the past few decades, this question has been hotly debated, and recently, there’s been an explosion of interest, research, and books written on the topic. (See The Light Eaters by Zoë Schlanger and The Nation of Plants by Stefano Mancuso as contributions from the past few years.)

Animism (from Latin: anima meaning “breath, spirit, life,” the belief that all things—objects, plants, animals, etc.—have agency and free will) has centuries-old tribal and religious roots. (See the Kalash in Pakistan, as an example). In addition, some scientists have argued that plants should be referred to as “intelligent,” because we can observe capabilities that we’d consider indications of intelligence in the animal kingdom. There is even an emerging field of research coined “plant neurobiology.”

However, we must be careful not to draw hasty equivalences. Simply lumping our observations about plants in with our understanding of human/nonhuman animal cognition, consciousness, and intelligence may distort both our definitions for these terms and our ability to appreciate what plants do very well.

How to tread the line between necessary investigation and fabulism? The answer may be found in avoiding excessive claims. We should focus instead on more modest, demonstrable accomplishment, while acknowledging our tendency toward anthropocentrism.

The Case for the Advocates

To get a sense for this complex landscape, let’s take a look at contemporary research into the evidence for plant intelligence. When researchers probe possible plant intelligence, they often draw on these known capabilities:

Plants exhibiting information processing & problem solving

An abundance of evidence shows that plants sense and adjust not only to the factors present in their environments, but also to changing situational variables. It’s reasonable to posit that they have a refined ability to process information and respond in ways that promote their own survival and development. Plants are adaptive organisms, constantly adjusting to shifts in light, water, gravity, temperature, soil composition, nutrients, toxins, predators, and more.

Source: Thoughtco.com

A commonly cited—and visually effective—illustration of this capacity is phototropism, the process whereby plants bend toward the strongest source of light. The process is enabled by the plant hormone auxin. Created and stored within tissues at the tip of a plant’s stem, it causes cells on the side of the plant facing away from the light to grow longer, thus promoting photosynthesis. In the face of fluctuation, plants are able to take in information and problem solve.

Plants are also sensitive to sound—they’re programmed to recognize resources, threats, and pollinators through vibrations. A 2017 study at the University of Western Australia found that the roots of pea plants are able to detect groundwater by sound and will grow toward pipes full of water making “watery noises,” even if the actual moisture remains completely inaccessible. It’s been posited that some plants can even make sound, though as far as we can determine, this is without intentionality. When placed inside a microphone-lined container, thirsty or damaged tomato and tobacco plants emit more noise than healthy plants, albeit at frequencies too high for the human ear.

Plants as information-storing and performance-improving

It is now understood that plants are not just reacting to incoming data; they are also storing and recalling information in a process analogous to memory. The Nasa poissoniana, a species of plant that grows in the Peruvian Andes, has been observed “raising their stamen, or fertilizing organs, shortly before a pollinator arrived, as if they could predict the future…could ‘remember’ the time intervals between bee visits, and anticipate the time their next pollinator was likely to arrive,” (Schlanger 2024). These flowers even seemed to be adjusting their behavior when bee-visit intervals shifted.

What’s perplexing about this type of adaptive recall behavior: It suggests some central structure, like a brain, for storing information. But as noted, plants do not have a structure homologous to the human/nonhuman-animal brain. Conceding this, some scientists nonetheless contend that the plant vascular system might be compared to the animal nervous system and might even contain similar molecular components.

Plants as communicators for individual and species survival

We have known for some time that trees communicate through their roots—sharing not only water and nutrients, but vital survival information as well about drought and disease. It is becoming clear that the plant world has methods of communicating in other ways and across longer distances. In the 1980s, chemist and zoologist David Rhoades discovered that trees in a Washington-state forest were sending warnings to other trees through chemicals in the air. When an invasive species of caterpillar plagued the forest, Rhoades observed that trees that the caterpillars hadn't reached yet had already altered the composition of their leaves to be poisonous as a means of protecting themselves. Given this finding, it seemed not only possible, but likely that trees with unconnected root systems were communicating in some manner about the danger. Rhoades’ conclusions were quickly dismissed as excessive by skeptics, but further research into plant communication has tended to support his conclusions.

Specifically: In 2002, two Cornell scientists—chemical ecologist Andre Kessler and doctoral student Michael Mueller—found that goldenrod plants can sense neighboring plants and adapt their responses when they become prey for herbivores like beetles. The goldenrod plants being eaten give off airborne chemicals (called volatile organic compounds, or VOCs) to communicate to the insect that the plant is not a good source of food; this information is also assimilated by close-by goldenrods, prompting them to grow faster and begin producing compounds that will help them defend themselves from beetles.

Similarly, corn plants can safeguard themselves from being eaten by caterpillars by emitting chemicals that attract parasitic wasps, who then lay their eggs inside the caterpillars—effectively killing them (Kolbert, 2024).

Kessler and Mueller propose to call this kind of behavior intelligent: The plant is problem solving based on information from the environment with the particular goal of survival. In a recently published report, they put it the following way: “the question is not whether plants express intelligent behavior but how they achieve it without a nervous system and what the ecological consequences of these behaviors entail," (2024).

This just in: plants with homologous chemical properties

It has recently been discovered that almost all plant species have neurotransmitters such as serotonin, dopamine, and glutamate—though the reason for this is still a subject for speculation. In any case, the presence of these homologous chemical messengers has strengthened the argument for plant intelligence.

A Skeptical Perspective

As much as the above findings are affecting, perhaps even persuasive, it’s important to keep this in mind: Demonstrating one quality, or even multiple qualities of intelligence does not necessarily mean an entity merits the descriptor intelligent. (We want to avoid calling mountains or oceans—tractors or vacuum cleaners— intelligent without convincing evidence!) And as noted, different researchers use different definitions for the term, making locking down an answer even more complicated. Plants are without-doubt complex organisms with wide-ranging capabilities. Astounding in their faculty and resilience, their modular design allows them to lose up to 90% of their physical form without dying—they are far hardier than anything in the human/nonhuman-animal world. However, when it comes to the question of whether plants ought to be dubbed intelligent, several important challenges should be noted.

Lack of homologous structures

As previously mentioned, the lack of a brain-like structure in plants makes it difficult to simply call them “intelligent”—at least without expanding the term to a breaking point. What scholars know about intelligence stems from research on human and non-human animals, all of which have some kind of brain or neural structure that we can study. In fact, one of the great mysteries of the plant world is the ability to store information without a brain or nervous system (at least in any sense that we ordinarily deploy those terms). Some plant researchers believe a plant’s vascular system is comparable to a central nervous system. And though some might see this as too large a leap, it gives us a clue as to where information may be located in the absence of a physical brain.

Simon Gilroy and graduate students conducting independent research © Bryce Richter/UW-Madison

Simon Gilroy’s lab in Madison, Wisconsin conducts experiments on a small, flowering plant called Arabidopsis. Researchers there discovered that when the midrib is “wounded” by scissors or a pair of tweezers, the whole plant responds. Though one might expect this type of wound to have a localized response, signals are experienced across the entire structure. In a paper published in Science, the team described: “Within minutes, an undamaged leaf can respond to the fate of a distant leaf,” (2018).

Some researchers have suggested that an entire plant could be considered a brain. Andre Kessler, the Cornell scientist studying goldenrod, claims that “there is no good evidence for any of the homologies with the nervous system, even though we clearly see electrical signaling in plants.” He finds an idea posed by mathematicians a century ago more compelling: that plants are comparable to bee hives—each cell acting like an individual bee. As he puts it: “What that means is, the brain in the plant is the entire plant without the need of central coordination,” (SciTech 2024).

But botanist Harriet Rix argues, “if…the entire plant could be regarded as a brain-like command center, there is a fundamental question about what remains to command, and what consciousness without separation might mean,” (Rix, 2024).

The lack of a discernible brain—as well as the lack of a conclusive answer as to where and how information is stored and processed in plants—makes it difficult simply to call them intelligent.

Response versus intentionality

Another complication: When it comes to plant behavior, we have no way of determining intentionality (which would imply awareness or some form of consciousness). Because intention cannot be ascertained, we must rely on assumption and hypothesis. It’s hard to say whether the responses we observe are because of “some form of awareness” or are more accurately described as automatic or reflexive. It’s important to note that our lack of certainty about intentionality is also an issue when interpreting animal behavior—though we may have more clues in that realm due to the homologies between animals and humans.

How far to stretch terminology

In his book Feeling & Knowing, Antonio Damasio makes a case for a difference between intelligence and explicit knowledge, writing, “…plants have a basic form of cognition and remarkable intelligence, but they do not have explicit knowledge concerning the things they do…sensing alone does not entitle an organism to mind or consciousness,” (Damasio, 2021). Delineations like this may be helpful, but they are also challenged by a real problem—the lack of clear, universally-accepted definitions for terms like intelligence, awareness, consciousness, and feelings. Author of The Light Eaters and staff writer at The Atlantic Zoë Schlanger would add to this list communication as a problematically ambiguous word. She comments that the definition of communication “slip-slides between the realms of philosophy and science, finding secure footing in neither.”

For the most part, these terms are most readily applied to humans and non-human animals—presumably because it’s difficult to separate our understanding and use of them from our inherent anthropocentrism. Some researchers even offer this suggestion: in order to truly appreciate the capabilities of plants, we may want to use entirely different terminologies. As an example, when queried about plant intelligence, Simon Gilroy strictly does not talk about plants having intelligence or consciousness. Instead, he uses the term biological agency to describe the ways in which plants are in active pursuit of their survival and thriving.

But in using these different terminologies, we risk losing the opportunity to make vital comparisons across the plant, animal, and human worlds. It may be most beneficial, perhaps even necessary, to bring plants into the intelligence-conversation in a measurable way, without jumping to far-fetched conclusions. This is what we seek to do in this essay.

III. Reviewing the Claims and Evidence for Forms of Computational Intelligence

In the preceding sections, we have considered the claims for intelligence(s)—capacity, potential, demonstration—across the animal and plant realms of life. We turn now the newest contender—computational programs and machinery.

We date the start of the psychological examination of human intelligence to the beginning of the 20th century, when French psychologist Alfred Binet created the first tests of intelligence. Much subsequent work in psychology has examined the concomitant claims for intelligence in primates, and, more recently, in other animal species—notably cetaceans and birds. Work on plants has been carried out chiefly by biologists and botanists—and as noted, for many, this work stretches the word “intelligence” too far.

Scientists at the Dartmouth Summer Research Project on Artificial Intelligence ©Margaret Minsky

We can date the start of the field of artificial intelligence to the 1950s. Indeed, many authorities trace its origin to a meeting held on a New Hampshire college campus in the summer 1956 with the intriguing title, “The Dartmouth Summer Research Project on Artificial Intelligence.” At this conference, researchers from several laboratories reported on computational advances: the programs that they had designed carried out mathematical and logical operations which, if carried out by human beings, might have been considered intelligent. But it was by no means clear that such computers could—then, and perhaps ever—pass the classical Turing Test: participate in conversation in a manner that we associate only with normal human beings; or win games of strategy in which human beings excel. 

The story of the next 75 years has often been told—including in Howard’s 1985 book, The Mind’s New Science: A History of the Cognitive Revolution. There were many surprising successes (e.g. programs that could consistently defeat adult experts in challenging board games) as well as frustrating disappointments (e.g. programs incapable of carrying out convincing, substantive conversations with human adults). But in the last several years, the dam has decisively been broken. If we can designate many animals as displaying various forms of intelligence; and if we can even consider the possibility of plant intelligence, it seems indefensible (if not perverse) to deny the term “intelligent” to various computational systems. Such a negative decision would be a judgment based on linguistic definition, rather than one based on actual performances and targeted comparisons.

IV. Stepping Back: A new way of conceptualizing intelligence(s)

Thanks to the devising of Large Language Models (LLM)—and in particular, the ever more versatile versions of ChatGPT—it is clear that computers can now converse and carry out analyses: ones which if done by human beings, would be considered highly intelligent, perhaps even at the genius level. Such systems can consistently win at challenging games (like chess and Go); compose convincing college essays and letters of recommendation; create plausible and, indeed, attractive works of music or art; make legal judgments or decisions to hire and fire as expertly as trained and highly skilled human beings; and even carry out medical diagnoses and therapeutic encounters at a high professional level. Moreover, these programs (if you will) constantly improve, with a speed that may be hard to fathom—let alone convincingly explain. 

There will always be resisters—and we have on occasion found ourselves in their ranks—who hesitate to call these performances intelligent in the traditional human sense. But these resistances appear less convincing with each passing year—and indeed, such exclusionary standards are so high and so precious that few human beings (indeed, few human cultures!) would qualify or bear comparison. Many commentators and observers muse, or worry, that in the not too distant future, it will be non-human digital computational systems that clearly merit the descriptor “intelligent”…while the rest of the animal (and perhaps plant!) world look increasingly “stupid,” “ignorant,” or “dumb.”

In focusing on LLMs, we do not intend to neglect or minimize the various kinds of robotic devices that carry out human-like tasks. But for present purposes, we will simply assume that these electronic-and-mechanical entities can and will make use of LLMs. As an example, we will assume that a robot that cleans one’s office, or plays chess, or administers psychotherapy, or acts as a companion or caretaker, is simply a physical manifestation of a computational system.

As we’ve done with reference to intelligence(s) purportedly exhibited in the animal kingdom, let’s consider how computational systems fare with respect to Howard’s list of multiple intelligences.

Without question, at present, the competence of such artificial, or artifactual, systems is most readily demonstrated in the areas classically probed by intelligence (IQ) tests. That is, LLM and other powerful systems clearly exhibit the signs of high logical-mathematical and high linguistic intelligence—they perform as well as or better than members of homo sapiens, even those thought to be savants!

Properly devised and trained, such programs can also demonstrate:

Musical intelligence: Recognizing pieces of music, creating credible new ones in distinctive styles, performing standard compositions with appropriate nuances;

Spatial intelligence: Mastering games like chess or Go, robots remembering and navigating complex terrains;

Bodily-kinesthetic intelligence: Again, navigating complex terrains, playing instruments, picking up and adroitly manipulating objects of various shapes and sizes, competing successfully in athletic events;

Naturalist intelligence: Recognizing and grouping members of discrete categories of living entities (plants, animals) as well as human-created entities (commercial products, ranging from thimbles to airplanes, from cereals to salads.

So far, computational systems emerge as multiply intelligent! But when it comes to the personal intelligences, one should be more cautious. To be sure, there’s already evidence that programs can engage skillfully in exercises that involve diplomacy, salesmanship, gamesmanship, therapeutic interactions—when deployed by human beings, these would be seen as manifestations of interpersonal intelligence. At least in a ChatGPT version, a program can respond like another person (even a person of specific traits and propensities)—recounting what has happened before, what one knows and what one doesn’t know, and how one is likely to react under different circumstances.

Yet, we are certainly not alone in questioning whether such responses have the same ontological status as responses by a “pinchable” or “punchable” human being—one reflecting on his or her or their own “felt” life experiences. Hence, the invocation of “intrapersonal intelligence” —an understanding of oneself—can be seen as a “category error.” An actor can feign ecstasy or misery, but we should not assume that the actor is actually undergoing these states of body or mind.

John Searle

Here we encounter manifestations of the classic Chinese Room Argument—posed over 40 years ago by philosopher John Searle. Searle envisions a room: therein, a person who knows only English is handed a very large set of questions and comments in Chinese characters (Pile 1). The person is given English instructions to respond using a matched set of cards in Chinese characters (Pile 2) that consists of reasonable responses to remarks emanating from Pile 1. Searle’s claim: it is a mistake to consider, to invoke, any form of understanding (let alone “intelligence”) here—the apparently conversing system is just matching Q and A, in a dumb, reflexive matter.

But many scholars have claimed that there is indeed understanding—though they have disagreed about the locus of the understanding. Some commentators have reduced human beings to simply being players in a universal language game; others have maintained that understanding is simply a way of describing what any entity does when it responds appropriately—no neurons, no intensions, no intentions, no hormones, no roots needed.

The “Symptoms” of Intelligence

Clearly, it would take much more thought—and many more pages—to sort through the concept and, indeed, the competing concepts of intelligence: how best to define it; how it is optimally deployed; and—among human, animate, live, mechanical, or digital entities—where to anoint and where to withhold this presumptively positive descriptor. We consign this task to the next generation of commentators and analysts—be they skins and bones, silicon chips, or some amalgam of physical neurons and digital neural nets.

We propose a different tack. We think it makes more sense to outline a set of criteria that one can apply to any system—human, animate, plant, computational, or some compound or amalgam thereof—and designate the ways in which any candidate qualifies or fails to qualify as intelligent. Of course, in doing so, we acknowledge that—in the present case—the proposers are human beings. Representatives of other cohorts or entities (including a properly-primed LLM) might take a different tack and reach a different conclusion.

Here we borrow gratefully from philosopher Nelson Goodman, who launched our research group, Harvard Project Zero, almost sixty years ago. In seeking to formulate what makes one entity or process “artistic/aesthetic” and another entity or process “non-artistic/non-aesthetic,” Goodman proposed a set of criteria that he dubbed “symptoms of the aesthetic.” When most of these were present, the candidate entity was deemed artistic (e.g. an epic poem, a carved statue); when few or none were present, the candidate was deemed non-artistic (e.g. a sales bill, a refuse bin). And when the scorecard was currently mixed, one awaited additional evidence. 

In our view, here are entities/criteria/judgments that should be applied to a candidate capacity. The “judge” (for now, a set of scholars like us) should designate whether the candidate fulfills the criteria easily; possibly; fails the criteria; or it is not possible to determine—either in principle or not at the present time. (The term “species” usually refers to animate entities; but we’ve extended it to include computational systems, in particular Large Language Models but also robots that carry out significant tasks).

1. Solving a problem—or seeking to solve a problem—that the (animate or digital) species may confront

2. Creating a product that may benefit the individual, the group, and/or the species

3. Communicating information—passing on skills or knowledge vital for survival or thriving (or that could threaten survival or thriving) for the individual, the group, or the species

4. Performance improving through practice/teaching/modeling/feedback.

5. Training/teaching/inducting other members of the species. This “symptom” is clearest when an elder is passing on knowledge or skills to a younger member; but at least conceptually, any trained program can be used to train any other program—either newer ones or older/less sophisticated ones!
   Note: Clearly criteria #3 and #5 are related; but #5 emphasizes passing on information to those entities that are expected to exist beyond the life span of the teacher/elder.

6. Exhibiting awareness or consciousness. Here our survey becomes tricky or problematic. Indeed, we might even suggest the need for a separate essay on “Who Owns Consciousness?”

Without question, non-living computational systems can be trained to testify to or feign consciousness. But to human beings—being conscious, and even more so, being conscious that one is conscious—constitutes the most impressive, but also the most elusive, of human capacities and indeed, of human intelligence. There is reason to believe that only in the last 10,000 years or so have members of our species achieved the kind of consciousness and self-consciousness that is nowadays taken for granted in most cultures. (Jaynes, 1976; see also Damasio, 2021).

Of course, one can also say that claiming one is intelligent should not be a criterion for being intelligent. We concede this point. But our claim here is that in ways that are instructive, many animals, possibly some plants, and certainly many computational systems merit the descriptor intelligent. They exhibit a reasonable sample of the aforementioned symptoms.

A Candidate That We Have Rejected for Now

Whether one member of a species (one rose, one rodent) is consistently better than another member. Indeed, this is one of the ways in which the notion of intelligence has been routinely applied to human beings—IQ tests purportedly document who is more intelligent according to explicit or implicit criteria. But except in cases where we (as adult scientists or experimenters) can control the environments of candidate entities, we cannot legitimately make such judgments. And we’d prefer not to be in the position of assuming such control over other forms of life. We might conclude that apes are more (or less) intelligent than baboons on one or another criterion. But we would hesitate to conclude that dog “Fala” is more (or less) intelligent than dog “Lafa.” By the same token, we might conclude that oak trees are more (or less) intelligent than rose bushes. But we’d refrain from comparing one tree or one bush with another.

To Summarize

Comparing different entities and types of entities on a set of criteria is always problematic. That said, it’s clear that many animal species—and most members of those species—readily exhibit several of the criteria for intelligence that we have set forth.

When one turns attention to plants, it’s more challenging to render judgments with confidence. One important consideration is the lack of homologies between plants and humans. Still, once one considers the behavior and activities of an entire species—rather than single members of that species—it is reasonable to attribute the descriptor “intelligent” to such cohorts. Certainly, the fact that goldenrod plants release airborne chemicals to warn nearby plants of predators is a discovery that deserves serious consideration.

Assessing the intelligence(s) of computational systems—and other non-animate entities—requires a different kind of stretch. Viewed from the stance of the behaviorist (contemporary followers of B.F. Skinner), these systems are clearly intelligent—indeed more intelligent than the species that (until this moment in time) created them! Moreover, these entities will undoubtedly continue to grow in intelligence, eventually exhibiting the full gamut of discernible human, animal, and plant intelligences…and possibly ones that we mere human mortals cannot even imagine.

Three Cautionary Notes

1. Even if computational systems claim to be conscious, we should treat such claims with caution. (Recall Searle’s Chinese room example.)

2. We are aware that, perhaps soon, certainly eventually, compound systems will be devised—perhaps Large Language Models will be merged with organic tissue, or human brains will receive implants of electronic neural nets. This has been dubbed “artificial general intelligence.”

3. Finally, whether animals, plants, computational system should be considered creative, ethical, compassionate, humane—those are challenges for another day…and for future “ownership” essays.

Closing Thought

To be sure, we have not provided a definitive answer to the question “Who Owns Intelligence?” We hope that we have usefully differentiated different facets of intelligence; and that we suggested some ways in which one can judge current candidates for that descriptor—human, animate, plant, computational, perhaps some day, compounds of these entities. And if we have contributed to a clarification of the question, and to ways to move forward in responding to it, our efforts have been amply rewarded.

---

We would like to thank Lynn Barendsen, Shelby Clark, Wendy Fischman, Kirsten McHugh, Danny Mucinskas, and Ellen Winner for their comments on an earlier draft.

SELECT REFERENCES

Binet, A., & Simon, T. (1905). New methods for the diagnosis of the intellectual level of subnormals. In H. H. Goddard (Ed.), Development of intelligence in children (the Binet-Simon Scale). Williams & Wilkins.

Charles, R. (2019). ‘The Overstory’ by Richard Powers wins Pulitzer Prize for fiction. The Washington Post. https://www.washingtonpost.com/entertainment/books/the-overstory-by-richard-powers-wins-pulitzer-prize-for-fiction/2019/04/15/b7701b28-5fa9-11e9-bfad-36a7eb36cb60_story.html

Cornell University. (2024). Are plants smarter than we think? Cornell scientists uncover form of intelligence in goldenrod plants. SciTechDaily. https://scitechdaily.com/are-plants-smarter-than-we-think-cornell-scientists-uncover-form-of-intelligence-in-goldenrod-plants/

Damasio, A.R. (2021). Feeling and Knowing: Making minds conscious. Pantheon Books.

De Waal, F. (2016). Are we smart enough to know how smart animals are? (First edition.). W. W. Norton & Company.

Gagliano, M. et al. (2017). Tuned in: plant roots use sound to locate water. Oecologia, 184(1), 151-160. DOI: 10.1007/s00442-017-3862-z

Gardner, H. (1983). Frames of mind. Basic Books.

Gardner, H. (1985). The mind’s new science. Basic Books.

Gardner, H. (1999) Who owns intelligence? The Atlantic, February 1999, 283, 2, 67-76.

Garnier, S., Gautrais, J., & Theraulaz, G. (2007). The biological principles of swarm intelligence. Swarm Intelligence, 1(1), 3–31.

Goleman, D. (1995). Emotional intelligence. Bantam Books, Inc.

Goodall, J. (1968). Behavior of free-living chimpanzees of the Gombe Stream area. Animal Behaviour Monographs, 161-311.

Goodman, N. (1981). Languages of art. Hackett Publishing Company, Inc.

Hage, S. R. (2024). Language evolution in primates. Science (American Association for the Advancement of Science), 385(6710), 713–714.

Harari, Y. N. (2024). Nexus: a brief history of information networks from the Stone Age to AI. Random House.

Jaynes, J. (1976). The origin of consciousness in the breakdown of the bicameral mind. Mariner Books.

Kessler, A., & Mueller, M. B. (2024). Induced resistance to herbivory and the intelligent plant. Plant Signaling & Behavior, 19(1). https://doi.org/10.1080/15592324.2024.2345985

Kessler, A. & Chautá, A. (2022). Metabolic integration of spectral and chemical cues mediating plant responses to competitors and herbivores. Plants, 11, 2768. https://doi.org/10.3390/plants11202768

Kolbert, E. (2024). Savvy in the grass. The New York Review. https://www.nybooks.com/articles/2024/10/03/savvy-in-the-grass-the-nation-of-plants/

Plants don’t have ears. But they can still detect sound. (2023). The Economist. Retrieved from: https://www.economist.com/science-and-technology/2023/09/06/plants-dont-have-ears-but-they-can-still-detect-sound

Powers, R. (2018). The Overstory. W.W. Norton & Company.

Rhoades, D. (1983). Chapter 4: Responses of alder and willow to attack by tent caterpillars and webworms: evidence for pheromonal sensitivity of willows. Plant Resistance to Insects. ACS Symposium Series, Vol. 208. The American Chemical Society.

Rix, H. (2024). Are plants intelligent? Book Review of The Light Eaters, by Zoë Schlanger. Times Literary Supplement. https://www.the-tls.co.uk/science-technology/natural-history/the-light-eaters-zoe-schlanger-book-review-harriet-rix

Schlanger, Z. (2024). The mysteries of plant ‘intelligence’. The Atlantic. https://www.theatlantic.com/magazine/archive/2024/06/plant-consciousness-intelligence-light-eaters/678207/

Searle, J. (1982). The myth of the computer. The New York Review. https://www.nybooks.com/articles/1982/04/29/the-myth-of-the-computer/

Shaw, J. (2024). Decoding the Deep: Project CETI’s pioneering effort to unlock the language of sperm whales. In Harvard Magazine, July-August, 2024.

Skinner, B. F. (1938). The behavior of organisms: an experimental analysis. Appleton-Century.

Sparavigna, A.C. (2014). The ten spheres of Al-Farabi: a medieval cosmology. International Journal of Sciences, 3, 34-39.

Sternberg, R.J. (1985) Beyond intelligence: A triarchic theory of human intelligence New York: Cambridge University Press.

Stupu, Andrei. Unpublished paper on moral intelligence, delivered at the Association for Moral Education in October 2024.

Toyota, M. et al. (2018). Glutamate triggers long-distance, calcium-based plant defense signaling. Science, 361, 1112-1115. DOI:10.1126/science.aat7744

Wright, L. (2022). The Elephant in the Courtroom. In The New Yorker, (Vol. 98, Number 3, pp. 44–44). Conde Nast Publications, Inc.

Multiple Intelligences in Fiji

Dr. Rosiana Lagi, the Deputy head of the School of Pacific Arts at The University of the South Pacific (USP), is a proponent for incorporating multiple intelligences theory into the education system in Fiji, an island country in Melanesia, part of Oceania in the South Pacific Ocean. In a recent interview with FBC News, Dr. Lagi describes why she believes an MI-based approach to education gives children the best opportunity to thrive and grow.

“It helps students be creative in their thoughts and in the things that they do,” Dr. Lagi says. “[They] have a high self-respect for themselves…because they know what they are doing is useful and important.”

Dr. Rosiana Lagi

In the interview, Dr. Lagi raises a few potential concerns she has for children receiving an education in Fiji—issues she has observed in her time as a leader in various academic spaces. “When our children are growing up,” she says, “they learn through play, they speak to objects, they speak to an imaginary audience…when it comes to school, these things are discouraged because we are content-driven.”

Dr. Lagi believes schools should incorporate exploration of students’ multiple intelligences into the curriculum. At USP, she encourages other teachers to build into their lesson plans diverse activities that use various intelligences and strengths. She highlights the arts—particularly singing and dancing—as engaging mediums to experiment with.

Learning of Dr. Lagi’s advocacy, we were reminded of the good work being done by MIS (Multiple Intelligence School), a private primary and secondary school based in Suva in the Fiji Islands. You can find more information about MIS’s efforts on their website’s welcome-page. 

It’s always exciting to discover new examples of MI theory’s widespread impact, and to hear that it’s being considered and applied in places far beyond the U.S. We are glad to read about Dr. Lagi’s championing of MI theory as a way to enhance early childhood education in Fiji. We hope that both her ideas and drive to improve student learning continue to find support.

The Trouble with MI "Quizzes"

Recently, an article was posted to Verywell Mind’s website offering visitors a fast-and-free quiz to “discover which type of multiple intelligence describes you the best.” The quiz asks about the quiztaker’s favorite subject in school, hobbies, favorite genres of TV, but also scenario-based questions, such as:

You’re sitting in the dentist’s office waiting for your appointment. How do you choose to pass the time?

Intrigued, I took the test myself, and after ten questions, was told my intelligence-type is verbal-linguistic intelligence. “You’re an expert at using your words, whether you’re writing or speaking,” my result-summary explained. “Verbal-linguistic folks like you tend to have solid memories, like to tell stories, and enjoy a good debate every now and then.”

While we have no doubt that this article and its author are well-intentioned and simply intrigued by the theory’s potential uses, Howard Gardner himself does not endorse MI quizzes or tests as a sound way to assess a person’s intelligence profile. It’s also important to note that if you see one of these assessments online claiming to be an “official” MI test, Gardner has never developed such a test, nor does he endorse any. The closest he’s come in the past is his role in the creation of Project Spectrum materials, which you can learn more about here.

             Branton Shearer

Many people have developed their own measures to test for multiple intelligences. The best-known instrument is the MIDAS. This test, developed by Branton Shearer (who may be reached at sbranton@kent.edu), has been administered to thousands of people all over the world.

Generally, Gardner’s trouble with tests that purport to assess a person’s intelligence is that most assess interest and preference rather than computational capacities, when only the latter indicates the strength of an intelligence. They also suffer from two major deficiencies:

1) They do not get at actual strengths of an intelligence – you would need performance tasks to determine how musically or spatially intelligent a person is.

2) They assume that the person has good intrapersonal intelligence and knows themself well. But many of us think we know ourselves better than we really do.

Many of these tests are not harmful in and of themselves. In fact, they may provide an interesting data point that leads to further investigation into our own strengths and weaknesses. And it’s understandable that many people, especially policymakers, desire an official instrument for measuring multiple intelligences. However, Gardner believes that optimally, intelligences should be assessed by more than one measure. For example, if individuals rate themselves on their intelligences using such a test, but also obtain ratings from those who know them well (family, friends, present and former teachers), the profile of intelligences would then be more reliable.

But the gold standard for assessing intelligences consists of performance measurements, where you must demonstrate your intelligence and not just testify to it. So, for example, one might:

• Assess interpersonal intelligences by observing how a person handles a conflict situation or motivates others to pursue a certain course of action.

• Assess spatial intelligence by seeing how quickly a person masters an unfamiliar geographical terrain.  

As for actual efforts to assess the range of intelligences in terms of performance, we know of only two examples: "Project Spectrum,” as mentioned above; and the “Explorama” at Danfoss Universe, a site at a Danish amusement-and-science park which allows visitors to profile their own intelligences. Howard briefly describes this in Multiple Intelligences: New Horizons. And in the book, Multiple Intelligences Around the World, there is a fuller description of the Explorama written by Charlotte Sahl-Madsen. 

                           Universe Park in Als, Denmark, where Explorama is located at Danfoss Universe

There is a disclaimer at the bottom of VeryWell Mind’s intelligences quiz results-summary, that reads, “Your results are not the end-all be-all…Consider your result as a fun way to think about your strengths and weaknesses, and the kinds of things you’re interested in.” We agree that these tests may give site visitors a good reason to look further into their own inclinations and abilities, but we also believe that the results of any such test should be taken with a grain of salt.  

Addressing Critics of MI Theory

Recently I received a letter from a colleague who was sympathetic to the theory of multiple intelligences, but was being hounded by individuals who believed that intelligence was singular and that it could only assessed by psychometric instruments. Here is what I wrote to the colleague.

***

Many thanks for your letter.  I always like to hear from those who have encountered my ideas and the ideas (including criticisms) of my colleagues. 

But, in candor, I have to say that your letter saddens me.  Over a forty-year period I have sought to make my ideas—their  sources, their claims, their implications, their limitations—clear. You can find the main points in my various books in education, on the MI Oasis website, which has posted dozens of blogs, and in two volumes of my collected papers, The Essential Howard Gardner on Mind and The Essential Howard Gardner on Education.

The critics whom you cite are fighting wars of the last century!  They are not open to new ideas, new ways of thinking, because their minds have already been made up...and appear to be calcified. 

Contrary to their claims: 

There are many kinds of science, many views of science.  Indeed, science changes with every decade—just think of the impact of microscopes, X-rays, cyclotrons, CRISPR, powerful computers, Large Language Machines, etc. And social science, a term I consider to be hyperbolic, is not the same as particle physics. 

The same goes for theory—many views of what a theory is. In history alone,  there are scores of theories about  history—the same in musicology or clinical  psychology. Importantly, there are many ways to test ideas and find out which ideas are worthwhile, which are worth pursuing and critiquing, which have educational implications, which do not. 

I cannot take seriously: the notion that intellect—and our research team is now studying animal, plant, and artificial 'intelligences'—can only be ascertained by a short paper and pencil (or computer-administrated) test.

Nor can I take seriously: a test that claims to determine one's intellect, one's potential, one's place in the world. 

Such notions could only be clutched tightly and retained indefinitely by ideologues. I speak, write, and address individuals  who have open minds (which is different from having intelligence, or multiple intelligences!). It is not worth trying to address individuals who have already closed their minds to any view of intelligence other than that developed over a century ago by psychometricians, some of whom were  open to having their minds changed. 

I hope that these brief remarks are of some help to you. If not, I am sorry. 

If you’re interested in my response to MI theory erroneously being labeled a “neuromyth,” see my article “Neuromyths: A Critical Consideration.”