Plant cognition after Darwin: historical and epistemological remarks

“Instinct is a modification of bodily structure (connected with locomotion.) no, for plants have instincts either to obtain a certain ends, & intellect is a modification of instinct – and unfolding & generalizing of the means by which an instinct is transmitted” (Darwin 1838–39, n. 48, my italics).

This may suggest that Darwin was just ready to attribute some “low level” cognitive powers to plants. However, regarding sensitiveness, he very clearly saw the major disanalogy with animals in the absence of nerves. As the Darwins put it in The Power of Movement in Plants:

The most striking resemblance is the localisation of their sensitiveness, and the transmission of an influence from the excited part to another which consequently moves. Yet plants do not of course possess nerves or a central nervous system (Darwin & Darwin, 1880, p. 572).

Given this difference, contemporary physiologists conceived an alternative way of addressing consciousness by looking for its molecular basis. In this regard, Darwin’s peers were sharply divided. Thomas Huxley (1869, p. 18) argued that “thoughts are the expression of molecular changes in the matter of life”. His view was that conscious processes supervene on physical processes at the molecular level, although this does not entail any explanation of consciousness. This agnosticism was close to Darwin’s own view. In the notebooks, he had written that “the mind is function of body” (Darwin 1838–39, n. 5), but in The Descent of Man (1871) he pointed out that this relation does not allow us to glimpse how mind evolved in organisms:

“In what manner the mental powers were first developed in the lowest organisms is as hopeless an inquiry as how life itself first originated. These are problems for the distant future, if they are ever to be solved by man” (Darwin, 1889, p. 66).

This agnostic view was strongly different from the view of other prominent naturalists. Alfred Wallace, for example, maintained that “either all matter is conscious or consciousness is, or pertains to, something distinct from matter and in the latter case its presence in material forms is a proof of conscious beings, outside and independent of, what we would term matter” (Wallace, 1870, p. 209). Ernst Haeckel, a prominent defender of Darwinism in Germany, straightforwardly sided with Fechner’s monism. Darwin had read the General Morphology of Organisms (Generelle Morphologie der Organismen), where Haeckel presents monism, the view that matter and spirit are inseparable in whole universe – which we call panpsychism –, as the “indispensable foundation of natural science” (Haeckel, 1866, II, pp. 441–442). In numerous articles and in the widely read book The Riddle of the Universe (Die Welträthsel, 1899), Haeckel spoke of “atom-souls” and “cell-souls”, thereby granting the psychic activity of plants.

“The plant-soul (phytopsyche) is, in our view, the summary of the entire psychic activity of the tissueforming, multicellular plant […]; it is, however, the subject of the most diverse opinions even at the present day. It was once customary to draw an essential distinction between the plant and the animal, on the ground that the latter had a “soul “ and the plant had none. However, an unprejudiced comparison of the irritability and movements of various higher plants and lower animals convinced many observers, even at the beginning of the century, that there must be a “soul” on both sides. Fechner, Leitgeb, and others strongly contended for the plant-soul. But a profounder knowledge of the subject was obtained when the similarity of the elementary structure of the plant and of the animal was proved by the cellular theory, and especially when the similarity of conduct of the active, living protoplasm in both was shown in the plasma theory of Max Schultze (1859)” (Haeckel, 1923, pp. 128–129).

Against this background, Darwin sided with a kind of agnosticism (see Smith, 1978). His view can be compared to the famous claim (“Ignorabimus”) formulated by the physiologist Emil du Bois-Reymond in the 1872 lecture “On the limits of natural science”. According to du-Bois Reymond, consciousness in its most basic form, i.e. “sensation” (which can be identified with today’s “phenomenal consciousness”) would remain forever unexplainable. To be sure, Du Bois-Reymond was also a staunch supporter of Darwin’s theory of evolution and, in the same speech, he praised The Descent of Man as a work which “forces upon him [the natural scientist] the theory that the soul came into being as the result, gradually attained, of certain material combinations, and that probably, like other heritable endowments that are of use to the individual in the struggle for life, it has risen and perfected itself up to its present state through a countless series of generations” (du Bois-Reymond, 1886, p. 127). In contemporary terms, du Bois-Reymond admitted the supervenience of mental processes on neurophysiological processes and he also granted that consciousness was a natural “fact”. However, he claimed that “not only is consciousness [i.e. sensation] inexplicable by its material conditions in the present status of science, which every one will readily admit, but […] even according to the nature of things, it never can be explained by these conditions” (Du Bois-Reymond Ibid., p. 117). From this perspective, Fechner’s “Plant-” and “World-souls” were not scientific, and Haeckel’s ideas on animated atoms followed the “the spirit of a false philosophy of nature” (1886, p. 123; p. 388, 412n8).8

Darwin’s view was less categorical: he admitted that the origin of consciousness in non-human organisms was outside the scope of present science. Thus, analogy – supported by the shared genealogy of organisms in general – was the only ground for extending consciousness beyond humans. Given the absence of a nervous system, the analogy with plants was merely grounded in a kind of “behavior”. This is precisely the point from where contemporary botanists have begun to reconsider the question.

The controversial status of plant cognition in contemporary botany is best understood starting from the publication of the popular book The Secret Life of Plants (Tompkins & Bird, 1973). In this bestseller, the authors made bold and scientifically unsupported claims about the feelings and intelligence of plants, including a musical preference for Beethoven and the capacity for mindreading. In response to early dismissive criticism of the latter claim, former CIA agent Cleve Baxter and IBM researcher Marcel Vogel replied that one had to develop empathy with plants to understand these facts. Following this episode, the scientific community developed a negative bias toward investigations into plant cognition.9 Nevertheless, the topic resurfaced with new research and promising results concerning plant sensation, communication, kin recognition, memory and learning.

The thesis of plant sensation has been recently explored by Baluska and Yokama (2021), who showed that plants are sensitive to the administration of anesthetics. This result is analogous to those that have been done with fish to prove the latter’s sentience, and similar investigations on pain markers have been conducted with insects.10 Overall, one implication of these studies is to bridge the gap between plants and animals.

Thus the transition from animal behavior and sentience to plants faces two problems: the first is justifying sentience beyond mammals, a topic that is itself debated (although arguments based on biological continuity and new experiments are leading to a widespread attribution of sentience beyond mammals: see Godfrey-Smith, 2021; Birch, 2024, Cf. below Sect. 6); the second is extending the analogy beyond animals. Regarding this second step, the question appears uncertain. On the one hand, it has been established that non-neural cells in both animals and plants can store information and influence biochemical responses. But these biochemical processes cannot be easily equated with animal’s sensation, since animal perception and response are adapted to their mobility and search for food (cf. Rix, 2025, pp. 33–34). Hence, some scholars talk of a peculiar “sense” in plants.

A similar point concerns plant communication. Several observations and experiments have shown that trees send chemical signals when harmed by caterpillars, fires or other physically damaging factors, thus sparking a protective reaction in other plants. Again, the controversy regards whether this can be considered as evidence of intentional communication (Schlänger, 2024, pp. 52–74). A connected, striking hypothesis concerns kin recognition. It has been argued that Douglas firs send nutrients to their seedlings and are capable of recognizing kin from non-kin, which could imply a kind a self-representation (Asay, Simard, Dudley 2020). Susanne Simard, among those who pioneered this line of inquiry, investigated trees to understand the way they “perceive and communicate with other plants, insects and mushrooms”, concluding that trees “cooperate” and particularly help their offspring (Simard, 2022). These views – popularized under the phrase “Wood Wide Web” – openly challenged the biological conception of life as governed by competition for survival. Simard’s insights stemmed from her discovery of the importance of biodiversity in the forestry practices. Once again, the evidence that trees connect with other tress through mycorrhizal networks is disputed in the scientific literature (Karst et al., 2023), and the anthropomorphisation of trees as communicating through the mycelium and caring for their offspring has been deemed inadequate (Jones et al., 2023. Cf.; Rix, 2025, pp. 128–130).

In a striking paper, Monica Gagliano and her co-authors reported an experiment with garden peas (Pisum sativum), aimed at testing whether plants could undergo a kind of Pavlovian conditioning. Pea shoots were set in a Y-maze structure made of pipes, and the experiment showed that two-thirds of the shoots grew in the direction of signals associated with light, indicated by a neutral cue (fans). According to the authors, this proved a “conditioned behavior”: the plants could learn and predict the direction of a light source. “This learned behaviour prevailed over innate phototropism” (Gagliano et al., 2016).

In a critical reply, Markel (2020) reported that she had replicated the experiment but found that only 50% of the plants grew in the direction of the neutral cue. The original authors replied that the experimental setup had been different (Gagliano et al., 2020). In a recent paper, Ponkshe et al. (2024) describe their attempts at replicating Gagliano’s experiment and address the methodological issues raised by Markel, suggesting ways to facilitate further investigations. (Meanwhile, evidence of plants keeping track of visits by pollinators has added more elements to the debate on plant learning. See Mittelbach et al., 2019).

Ponkshe and her co-authors conclude that “ultimately, we need a comprehensive, comparative psychology of learning across the tree of life, respecting phenotypic comparisons to truly understand behavior’s diverse forms”. As this conclusion suggests, the epistemological issue – and controversy – at stake is broader. Gagliano’s experiment might merely prove mechanisms of associative memory in plants, without entailing any claim about sensitiveness and consciousness, but its ultimate scope is to foster investigations into further plant cognitive powers. In her 2018 book Thus Spoke the Plant, Gagliano maintains that plants think, feel and communicate with other plants – and even with humans – grounding this view in a mix of experimental and subjective experiential evidence.

Overall, these findings depend on a combination of new evidence and new interpretations. Their use of notions of intelligence, memory, behavior, and self, drawn from the study of animals and extended to plants, aims to question the traditional distinction among living beings. This raises two distinct issues, which are clearly illustrated by the debate over Gagliano’s results. First, there is the standard scientific need for experimental replication; second, the justification of conceptual innovation. In this latter case, the opportunity for analogy with animals has led to major controversy among scientists. A key moment has been the proposal, put forward by a group of researchers, of a new field of “plant neurobiology”. This provocative label was intended to highlight the possibility of a different realizations of cognitive functions in organisms lacking nervous systems, and to counter what they saw as a kind of “self-censorship” in contemporary science (Brenner et al., 2006). Following this trend, the discipline of cognitive ecology focuses on the functional role of learning in nature, with the working hypothesis that “brains and neurons […] may not be a necessary requirement for learning” (Gagliano et al., 2014).

In a critical article, Lincoln Taiz and his co-authors noted that “regulatory interactions are occurring throughout the plant”, rather than being concentrated in the root tip. Therefore, “we could regard the entire plant as a brain-like command center, but then the brain metaphor would lose whatever heuristic value it was originally supposed to have”. The authors concluded that it is “extremely unlikely that plants, lacking any anatomical structures remotely comparable to the complexity of the threshold brain, possess consciousness” (Taiz et al., 2019).11

In light of this survey, the legacy of Darwin’s root-brain hypothesis – evoked in the paper on “plant neurobiology” cited above – remains controversial. The question is whether Darwin’s analogy, dismissed by 19th-century botanists, appears more plausible in a contemporary context. Certainly, the different theoretical framework may play a decisive role. As we have seen, revolutionary views on plant cognition have been inspired by theories of cognitive science such as the multiple realizability of cognitive states. Moreover, recent investigations on animal minds may have further encouraged the idea that traditional ontological boundaries can be broken. The significance of this new context will be explored in the final section.

The conceptual dimension of the controversy over plant cognition has been finely characterized by Daniel Chamowitz (2018). Given the evidence that “plants integrate many signals to adapt to their environment and increase their fitness”, Chamowitz asks: “Is this a proof of intelligence? It depends on the meaning of the word”. Those who attribute intelligence to plants define this capacity as a form of problem-solving displayed over an individual organism’s lifetime. Thus, Anthony Trewavas proposed defining plant intelligence as “adaptively variable growth and development during the lifetime of the individual” (Trewavas, 2003, p. 1). Advocates of plant cognition connect this kind of definition to the thesis that neural networks are not a necessary condition for cognitive states. After the controversial term of “neurobiology” was replaced with “signaling and behavior”, Paco Calvo restated the general claim of this approach, which.

“aims to unearth that plants perceive and act in an integrated and purposeful manner, and how they do it. The rationale that underlies this effort is the idea that intelligent, flexible behavior requires coordination among diverse plant structures. […] the field [thus] proposes an interdisciplinary and integrated view of plant signaling and adaptive behavior in order to study plant intelligence” (Calvo, 2016, pp. 1323, 1326).

In the critical paper cited above, Taiz and his co-authors (2019) protested against the resurgent “anthropomorphism” of those who talk of learning and feeling of plants. Calvo and Trewavas (2020) replied: “This is surely circular reasoning”. This exchange clearly shows the epistemological crux of the debate on plant cognition. Talk of plant “intelligence”, “learning”, “feeling” and “self-recognition” is based on an analogy with humans. The question is: How far can we meaningfully extend the analogy? A possible perspective is to reduce the issue to a matter of terminology. For example, Calvo attempted to settle the issue as follows:

“It goes without saying that plants do not share “neurons” with animals, or exhibit animal “intelligence”. If the reader wishes to keep those terms for animals exclusively, so be it. Plant neurobiology interprets plants as information-processing networks with individual cells as computational building blocks” (Calvo Garzón, 2007).

Yet this emphasis on function and computation does not entirely solve the problem. Because we use psychological terms to define human cognitive functions, the question is whether it is adequate to attribute a different kind or degree of the same functions to plants.

A similar problem has arisen in cognitive ethology. Given that definitions must start from human experience, the issue is whether it makes sense to extend the analogy to other species. In his investigation of the roots of moral capacities in primates, Frans De Waal (2016) pointed out that, in the light of evolutionary biology, “anthropomorphism” must be tempered without lapsing into the opposite mistake of “anthropodenial”, the claim of absolute human exceptionalism. The search for precedents of human morality in primates was motivated by shared biology and was, in fact, a development of Darwin’s ideas.12

Peter Godfrey-Smith has made a similar reasoning with respect to consciousness, starting from a critique of behaviorism:

This animal is only doing things that we know can be done unconsciously. It would then feel like nothing to be that animal. Next, this conclusion might be applied to various actual animals, whose abilities are similar to the hypothetical one. This would not be a good argument, however. The experiments the argument is based on were all done with conscious human beings, even though the people were not conscious of everything going on inside them. From these experiments we are learning that, when you are a conscious human being, a surprising amount can be done deep in the background. That does not show that everything could be in the background at once. For any normal and wakeful human being, there is something it’s like to be that person, even if much is being done behind the scenes. The same may apply to a great many animals as they make their way through the world […] The mistake that can then be made is to say: this (as studied in humans) is what experience is, and if you don’t have this going on, you can’t have experience. The alternative is that if your brain is different from ours, your experience is different, but not absent (Godfrey-Smith, 2021, ch 10).

Godfrey-Smith’s search for different kinds of minds was rooted in Darwinian ideas, particularly the view of genealogical continuity among living beings. The thesis of multiple realizability of mental states is a significant post-Darwinian element here, with the octopus serving as a model of an animal which is both very intelligent and very different from humans. This showing how the meaning of cognitive functions changes with the extension of the analogy beyond species that are closely related to us on the evolutionary tree. In his materialist perspective, Godfrey-Smith adds an important dimension to the logical problem of analogy: different bodily structures determine, to varying degrees, differences in mental capacities.

Given the difference between animals and plants – from the cellular level to the architecture of the body – Godfrey-Smith treats plant cognition as peculiarly distinct from all animal cases. First, he argues that nervous systems operate differently from plant biochemical signaling systems: “what nervous systems do is achieve specific kinds of cell–cell interaction—interactions that are fast, targeted, and directional” (Godfrey-Smith, 2016), while signaling in plants is massive and does not happen with fast cell–cell interactions. Second, it is difficult to define plant individuals and, hence, plant subjects:

“A plant is in some ways like a community rather than a single individual. Within a community there can be signaling and coordination; you and I might communicate closely, and still be two. Social interaction does not merge our subjectivities […].