Understanding interdisciplinary perspectives of plant intelligence: Is it a matter of science, language, or subjectivity?

Blue light was the “food” source, a fan was the neutral cue, and the response was growth toward the stimuli. Results from Gagliano et al. [16] demonstrated that the pea plants were able to associate the wind with where the light was going to be. In the absence of blue light, the plant was still able to associate that the direction of the blowing fan would determine where the light source would be. Not only did the green pea plants have to respond to two external stimuli, but it also had to internally and globally process that those two stimuli were associated. This type of learned behavior can be seen as an expression of an anticipatory behavior that required storing and processing information and globally responding [16]16.59.91422.321.515.71213.7(3) Salt avoidance in Arabidopsis thaliana (L.) Heynh Roots need to make the best overall decision. Apart from growing toward gravity, water, and nutrients, they must sense sources of stress such as salinity. Roots of Arabidopsis thaliana (L.) Heynh. Navigate belowground to avoid high saline conditions. The root apparatus possesses a sophisticated sensory and communication system that allows the plant to respond flexibly. Roots must integrate these and many other cues so as to grow one way or another under constantly changing conditions [23]23.321.635.319.81162.5(4) Host preference in parasitic plants The parasite dodder (Cuscuta pentagona Engelm.) obtains its nutrients from the shoots of the host plant it climbs onto. By sensing the airborne volatile compounds that potential hosts emit, they are able to choose the more nutritious ones. In the vicinity of a tomato and a wheat plant, the dodder can tell them apart and grow toward the more nutritious tomato exemplar [25]22.419.837.919.81162.5(5) Mycoheterotrophic parasitic plant (mere behavior) Mycoheterotrophic plants cannot do photosynthesis.

They obtain all their water and nutrients from the fungi they parasite providing nothing in return. Many of them mimic fungi varieties other than their host, remaining belowground and becoming visible only when they flower. Their metabolic cost in doing so is minimal [54]32.22033.913.91152.3(6) Secondary metabolite manipulations of nectar in plant–ant relationships Many plants secrete extrafloral nectaries as food for ants in return for their protection. Plants need to attract the right kind of ant partner and actively maintain their quality protection. The nectar is custom-modified to specific ants. Too little nectar can discourage ants away while providing too much nectar can lower the quality of protection. These plants have to sense the presence of different ants, monitor and modify their activity accordingly in order to get the best benefits [19]21.718.337.422.61152.6(7) Resource exploitation by roots In one study, roots belonging to the same pea plant (Pisum sativum L.) were exposed to two patches of nutrient: one patch with more amount of nutrients but whose concentration remained constant, and a second patch with less amount of nutrients, but whose concentration would increase throughout the experiment. The pea plants grew more roots to the patch that currently had less amount of nutrients, but would have more in the future due to the steady increase [26]25.216.541.716.51152.5(8) Plant defense turning herbivores into cannibals (mere behavior) Some tomato plants (Solanum lycopersicum L.) are able to defend themselves against herbivory by releasing certain chemicals (methyl jasmonate).

Researchers observed that spraying the tomato leaves with methyl jasmonate promoted cannibalism: munching caterpillars would prefer to change diet and eat other fellow caterpillars instead [55]33.316.72821.91142.4(9) Numerical sensitivity and short-term memory: Venus fly trap, Dionaea muscipula J.Ellis The Venus flytrap, Dionaea muscipula J.Ellis, snaps shut when it gets mechanically stimulated twice. After the first stimulation, the second needs to occur within 20 s of the first to snap shut. If it does not, then it resets. While counting to two may seem less impressive, counting to five is more difficult to do. Before digestive enzymes are secreted, the Venus fly trap needs to keep counting the numbers of mechanical stimulation until it reaches five. It is the plant’s extra security mechanism. Dionaea spp. are able to count, store information as a form of short-term memory, and repeat the process [14]27.213.14023.41142.7aResponse scale 1 = not intelligent to 4 (or 6) = very intelligentbMean Likert scores higher or lower than the expected mean (3.5 for examples 1 and 2 and 2.5 for the rest) indicate pro-plant intelligence views and anti-plant intelligence views, respectivelyNumbers in bold indicate mean Likert scores lower than the expected mean (i.e., anti-plant intelligence views)

Table 2 reports respondent agreement or disagreement with the individual statements as well as an aggregate opinion on plant intelligence by combining all plant behavioral examples that were indicative of plant intelligence. Statements 3 and 5 had the largest Cramer's V value (0.48 and 0.47, respectively), indicating that concepts related to plant communication and plant response to sound have a relatively stronger association with views on plant intelligence. The remaining Cramer’s V values had a relatively poor strength of association, suggesting that concepts less specific to human–plant interactions are poorly associated with individual opinions of plant intelligence.Table 2The number of respondents who agreed or disagreed with the statements and their aggregate opinion on plant intelligence measured by combining all plant behavioral examples indicative of plant intelligence. All statistically significant results had statistically significant pairwise comparisons (two proportion Z test)PerspectivesCombined opinion of plant intelligenceChi-square valueCramer's Vp valuenReferencesIntelligentNot intelligent1. A plant or tree has spoken to me beforeAgree1125.5970.230.018108[46]Disagree43522. I believe that hallucinogenic drugs reveal inner truths of an individual and other truths of the world. These “hallucinations” are not irrelevantAgree2597.1430.270.00894[56]Disagree28373. I believe singing or talking to plants doesn't have any effect on the plantAgree123422.8090.48< 0.001100[57, 58]Disagree41134. I don't believe in the significance of dreams; they do not hold any factual truth about the worldAgree15278.7190.290.003103[42, 46]Disagree41205. I believe that the Flora kingdom can communicate with us, either through dreams, visionary illustration and/or other means of communicationAgree35723.3430.47< 0.001105[59, 60]Disagree21426.

I believe that plants are teachersAgree40234.5370.210.033103[46, 60]Disagree16247. I would consider myself spiritual. By spiritual, it does not necessarily conform to a labeled religionAgree36167.1430.270.005107[49, 51]Disagree22338. I don't believe in the existence of a spiritual, unseen dimensionAgree31175.7250.240.017100[49, 51]Disagree21339. I practice mindful meditation when I can. I believe its practices allow me to connect with my higher selfAgree344411.4780.320.003115[49, 51]Disagree25610. I incorporate Eastern philosophy into my way of living such as Taoism, Buddhism and HinduismAgree1675.5970.160.033111–Disagree414711. I am a person who listens and follows what my heart says relatively more than my mindAgree21132.465–0.2999[61]Disagree273412. I feel strongly and personally connected to natureAgree57453.276–0.19117[5]Disagree37

Ethnobiologists (n = 15) had a significantly higher PP index (reflecting acceptance of non-Western ontological knowledge and beliefs) than non-ethnobiologists (Kruskal–Wallis chi-squared = 12.8, df = 1, p value = 0.0004, n = 66). A higher PP index was also significantly related to those who have a higher frequency of collaboration outside the biological sciences (Kruskal–Wallis chi-squared = 12.238, df = 1, p value = 0.0004, n = 56) and within the humanities (Kruskal–Wallis chi-squared = 4.368, df = 1, p value = 0.037, n = 53).

Seventy-six percent of respondents scored the animal behavioral examples of camouflaging in a cephalopod and climbing behavior of macaques as intelligent. Others had either mixed opinions (20%) (i.e., intelligent in one example and not intelligent in the other example), or did not think these behaviors were intelligent (4%). More respondents considered the Corydalis hemidicentra camouflaging behavior and Syngonium spp. foraging behavior as “not intelligent” (Appendix Fig. 2). Approximately 32% of respondents shifted their opinion from “this is an intelligent animal behavior” to “this is not an intelligent plant behavior.” Internal consistency (Cronbach’s alpha) of answers was 0.8.

The mean Likert scores for the camouflaging and foraging behavior in animals were 4.2 and 4.7, respectively, indicating pro-intelligence views. Furthermore, the corresponding plant camouflaging and foraging behavior had a mean Likert score of 2.5 and 2.6, respectively, indicating a marginal (or neutral) view of intelligence (Table 3).Table 3A list of behavioral examples provided in the questionnaire where respondents were asked to rank on a scale from one to four (or six) whether they considered the presented behavior to be intelligent or not (1 = not intelligent to 4 or 6 = very intelligent)Behavioral Examplesa% distributionnMeanb123456Animal behaviorCamouflaging: Cephalopod species (squids, octopus and cuttlefish) are known to be very intelligent invertebrates. The octopus have evolved an effective and impressive camouflaging ability that allows them to manipulate and exploit their surroundings to hide from predators and hide from prey. This intelligent behavior requires the ability to acquire external information, process complex information, and flexibly adapt in order to successfully achieve its goal [64]63.411.221.631.925.91164.2Foraging: The omnivorous stump-tailed macaques, Macaca arctoides, spends half of its day foraging and feeding. M. arctoides usually travel on the ground, but do climb trees either to forage for food resources or to go to sleep. They travel to areas where food is abundant, by remembering the exact locations and by following the changes in seasonal patterns of the vegetation. They can travel from tree to tree when they forage, avoid other territories belonging to other groups, and descend to continue to forage in other areas when food resources have been eaten up. Many climbing species, much like M. arctoides, forage for food in an adaptable, meaningful and flexible manner.

They know where resources are, avoid competition, and anticipate where food resources will be. They respond to these changes to ultimately maximize their fitness [65]0.92.52.51738.1391184.7Plant behaviorCamouflaging: Some plant species (Corydalis hemidicentra Hand.-Mazz) can camouflage like animals with the same known behaviors: (1) Background matching—blending with the colors of shapes of the habitat where they live; (2) Disruptive coloration—markings that create the appearance of false edges and boundaries, making it harder to see the true outline; (3) Masquerade—looking like something else; usually something a predator might ignore, such a stone or twig. Examples include living stones, some cacti, passion vines and mistletoes; and (4) Decoration—accumulating material from the environment. For example, some coastal and dune plants get covered by sand because of their sticky glandular trichomes, making them less conspicuous. This intelligent behavior requires the ability to acquire external information, process complex information, and flexibly adapt in order to successfully achieve its goal [66]18.930.630.619.81112.5Foraging: Climbing plants, such as tropical species from the Araceae family, Syngonium spp., forages for light by reaching the top of its host tree and descends to the ground while searching for another host tree. Its morphological features also progressively changes as it climbs, growing thicker filiform stems (tendrils) with larger and thicker leaves at the top. The tendril explores to find other tree hosts by extending and descending downwards, having thinner tendrils and smaller leaves. Syngonium spp. can therefore switch between mobile and sessile behavior. It has the capacity to decide when to climb and when to descend, by flexibly changing its growth direction given exogenous changes.

This light-foraging behavior is adaptable and meaningful [67]13.529.738.7181112.6aResponse scale 1 = not intelligent to 4 (or 6) = very intelligentbMean Likert scores higher or lower than the expected mean (3.5 for examples 1 and 2 and 2.5 for the rest) indicate pro-plant intelligence views and anti-plant intelligence views, respectively

One of the questions ethnobiologists were asked was “Have local people you worked with ever mentioned in their own way that plants are intelligent? If so, how?”. Fourteen respondents shared their field experiences. The main concepts ethnobiologists had recorded from local community members were: (1) plants have emotions, (2) plants can communicate with other plants and animals, (3) plants are conscious beings, (4) plants are considered kin, (5) plants are divine beings and have spirit(s), (6) plants need respect and require offerings, and (7) people can communicate with plants to learn from them and collect natural resources.

Subjective awareness was defined as “involving the ability to be aware, aware of one’s own body parts and existence.” To the question “Do you think plants can be subjectively aware?,” 42 participants responded with a dichotomous “yes” (n = 17) or “no” (n = 25). In addition, 77 responses pertained to seven themes (number of participants given in brackets): It depends on the definition (11); uncertain (28); yes, in a non-anthropocentric way (8); yes, as an intrinsic ability (11); no, there are no neuronal networks or thought patterns involved (10); open to the possibility (4); this cannot be proven or disproven (5).

Several respondents explained why they considered the examples intelligent behavior, despite the open question asking the contrary (i.e., why is this example not considered an intelligent behavior?). Explanations can be summarized under the following themes: (1) brains are not necessary for intelligence, (2) adaptation to the environment to maximize fitness is intelligent, (3) intentional behavior for survival is intelligent, (4) successful strategies and adaptable behavior require intelligence, and lastly (5) it is an intrinsic trait of a plant. They agreed that learned behavior, having memory, and the ability to make associations are a form of intelligence. In contrast, other respondents stated that basic learning, having memory, and the ability to make associations are insufficient to be qualified as a form of intelligence. Thus, explanations were diverse and showed divergent views. Table 4 gives an overview of all thematic explanations for each of the behavioral examples. Frequently, explanations included reasons that implied mechanical processes exhibited without any “intention” or “awareness” from a plant.

Another common theme of explanations was that the adaptive behaviors made by plants are fixed and controlled by genes, an evolutionary power that is not related to individual decision-making.Table 4A summary of the main co-occurring thematic explanations of individual responses for each of the plant behavioral examples asked in the questionnaireCo-occuring thematic explanationsHabituation in Mimosa pudica LConditioned learning in Pisum sativum LSalt avoidance in Arabidopsis thaliana (L.) HeynhHost preference in Cuscuta pentagona EngelmNectar in plant–ant relationshipsResource exploitation by roots of Pisum sativum LCounting and memory in the Venus fly trapMycoheterotrophic plantsTomato plant turns herbivore into cannibalsUncertain–52311322Response to external stimuli4472121––Depends on definition432211–11Physiological response6332–12–1Automated response418973553Evolutionary process1–2617382617No conscious thought involved318653663Chemical response––15–1––5Threshold response––––––51–Tropism––43–2–––Expresses concern about the study–46–24–––Case-specific themesSaving energyNot considered countingMimicry/parasitism is not intelligentResponse not directed by plant3547Total number of respondents534549474242435752The numbers represent the occurrence of one theme being mentioned per behavioral example. Case-specific explanations were themes unique to a behavioral example. A theme was only produced if it was mentioned more than three times in any of the behavioral examples. For the definition board where all themes were explained, see appendix Table 6. The last two behavioral examples on the right were behaviors not intended as indicators of plant intelligence but, instead, of mere adaptation

A small number of participants expressed confusion about the definition of intelligence or stated that intelligence is a semantic issue. Some people showed expressions of conflict and uncertainty by explicitly stating that they could not provide an opinionated response while some also questioned or reserved the meaning of intelligence exclusively for humans or humans and other animals. In addition, some were uncertain what constitutes intelligence (and the levels of intelligence) and how it can be theoretically applied to plants.

The explanations summarized in Table 4 demonstrated that the thought patterns of plant scientists can be divided into three main groups. One subgroup paid more attention to the mechanistic properties as a determinant of intelligent behavior. A second subgroup was concerned less with the mechanical processes, but interpreted these behaviors as controlled by genes which, due to evolutionary processes, determined beneficial adaptive behaviors. A third subgroup rather relied on what the behaviors represented as a whole, that is whether the observed behaviors conceptually indicated intelligence or not.

The subtle but crucial variations that differentiate mere plant behaviors from intelligent behavior were not noticed by most respondents. Such mere behaviors are responses that are fixed, i.e., genetically determined, and incapable of long-term change such as the example of mycoheterotrophic plants in Table 1. One possible explanation for this could be that a considerable degree of familiarity with the concept of plant intelligence is required to detect these subtleties. We tried to make up for this by introducing the concept of plant intelligence briefly prior to the start of the questionnaire. However, it likely requires training and experience to interpret and compare different types of plant behaviors. Regardless, our results show that there exist significant gaps in knowledge and communication about plant intelligence. These gaps are caused by uncertainties about what entails intelligent plant behavior and how it can be distinguished from other behaviors. Our results highlight the complexity and diversity of opinions on plant intelligence and show that an individual view of plant intelligence depends on several factors. As a result, based on our findings supplemented with the literature on factors contributing to the lack of consensus on plant intelligence [33], we developed a conceptual framework (Fig. 1) consisting of three key components and their factors that influence a scientist’s perspectives. This conceptual framework aims to explain why variations in perspectives of plant intelligence occur and highlights the gaps in knowledge and communication observed in our study.Fig. 1A conceptual framework on the components that can influence a researcher’s perspective of plant intelligence

Central to the conceptual framework are factors that contribute to the formation of a person’s identity, namely culture, environment, and beliefs. Culture plays a key role in valuation systems and decision-making processes [68, 69]. Interactions between humans and plants have significantly governed cultures and the surrounding landscape (e.g., [70]). Cultures have been conceived and continuously shaped by the various ecosystem services of plants (and nature as a whole). Culture contributes to the perceptual lens by which natural phenomena (in this case, plant behaviors) are observed and interpreted. Furthermore, the human–nature relationship is reflected through relational processes, for example, one’s sense of place, value systems, and connections to nature [71, 72], and lived experiences between people and the environment [4]. Culture and environment are therefore tightly interlinked and mutually shape each other [73]. Finally, an individual’s belief system is inevitably tightly interlinked with culture and the environment. A belief system is the philosophical foundation of how one decides to act within and relate to the environment [72]. As observed in this study, a belief system that either associates or disassociates with belief in the agency of plants, either implicitly or explicitly, reflects a scientist’s perspectives of plant intelligence, independently from existing scientific evidence. Our findings conform to claims made by other authors (e.g., [74]) that disagreements and non-recognition of plants’ behavioral ability had less to do with scientific evidence but more with personal belief systems, regardless of the evidence provided.

However, we acknowledge that plant intelligence is still a relatively new concept with minimal supporting experimental studies—e.g., independent replication of Gagliano and coworkers’ pea plant associative learning experiments is still lacking [75]; although see [76, 77]—and, evidently, will result in disagreements and miscommunication in the interpretations of various plant responses.

Both the environment and belief systems are embedded within a broader cultural dimension and constitute one’s worldview [78]. Here, a worldview is described as an intuitive view, a set of values, and belief system concerning one’s conception of life [78]. These three factors (culture, environment, and belief system) take part in the formation of a plant scientist’s professional identity, the knowledge system they hold onto, and interpretations of ecological phenomena. Therefore, interpretations of plant behaviors evidently reflect a scientist’s subjective opinions about plant intelligence, either implicitly or explicitly.

Respondents came from a variety of professional backgrounds, most of which do not require the involvement of other forms of knowledge (such as TK) in their work. However, involvement with TK systems and multidisciplinary collaborations may provide a complimentary and inspirational lens for scientific investigations and is even more essential when decision-making in resource management is involved [47, 48, 79–82]. In plant-related studies, it expands the ways scientists understand different plant behaviors more holistically, taking into consideration both mechanistic properties and the overall representation of ecological phenomena of an observed plant behavior. Our findings showed that plant scientists who more frequently included TK in their discipline agreed more with the idea of plant intelligence.

Interestingly, one study showed that social scientists have more of an organismic worldview (more holistic, interactive with, and integrated into social environments), while natural scientists have a more mechanistic worldview (more objective, reductionistic, and separate from social environments) [83]. Authors in [84] discussed how social and biophysical sciences have different levels of awareness about the relationship of cultural worldviews shaping scientific theory and practices. Our study provides a similar finding where two groups of plant scientists from different backgrounds (either ethnobiologists or non-ethnobiologists) also differed in their ontological views of nature and plants, especially with specific ontologies that are associated with traditional knowledge of plants. Plant scientists who included sociological aspects in their work and collaborated more frequently outside their disciplinary field tended to associate more positively with non-Western knowledge systems than sensu stricto plant scientists. These associations therefore take part in influencing a plant scientist’s perceptions of plant intelligence.

The language and meaning that is used to describe plant behaviors affects the way information about plant intelligence is being communicated and understood. Consequently, the majority of arguments and disagreements can largely be attributed to semantic issues and not the phenomena of plant behaviors. Some plant scientists may take the terminologies describing plant behavior literally when words are used as metaphors, such as the phrase “plant neurobiology” [85, 86] or metaphorically when words were meant to be taken literally, such as the phrase “plant intelligence” [87].

There are multiple definitions for “intelligence” without a current consensus [88]. This poses a problem where disputes will be mostly about the meaning of plant intelligence and determining what definition and theoretical framework of intelligence can be best applied to plants. Another issue is whether there should be a unified framework for intelligence, or whether different frameworks for intelligence should be allowed to exist [6, 89]. Different theoretical frameworks of intelligence will use their own terminologies to discuss their theories, approaches, and discoveries. Definitions (and theoretical frameworks) are also not set in stone and are subject to ongoing revision. Even when definition issues have been resolved and all parties have the same understanding of the meaning of intelligence and settle on the same theoretical framework, differences in understanding the same phenomena will still exist. This was also observed in our study, given that the definition of intelligence was provided prior to the questionnaire. According to respondents, the plant behaviors provided in Table 1, such as complex decision-making, anticipation, counting, and adaptability, were interpreted differently with explanations in favor or against intelligent behavior.

Problems with semantics and personal subjectivity are also common in other biological sciences, exemplified by the definition and diversity of perspectives about “invasions” in biological studies [89]. Therefore, open discussions are needed so that different subjective viewpoints can be shared, an end goal known as intersubjectivity. For example, the author in [90] discussed how achieving intersubjectivity can act as an intermediary by actively coordinating participants who disagree with each other to express their subjective opinions. As a result, this can allow communicated viewpoints to be less subjective.

The scientific paradigm of plant intelligence can benefit from the integration of ethno- and other social sciences, because of their relationship with multiple worldviews. The concept of plant intelligence has been mentioned in various socially oriented disciplines, e.g., philosophy [7, 91, 92], ethnobotany [93], and anthropology [94, 95]. Meanwhile, to our knowledge, only a few literature references in the biological sciences incorporate the ontological features of plant intelligence (e.g., [96, 97]). The results of this study make the case for facilitating an interdisciplinary dialogue within various plant-related disciplines among those that argue for and against plant intelligence, so that these scientists may take away multiple perspectives and adopt them during their scientific practice [47]. Researchers will be able to express their personal viewpoints and explain their unique interpretations of plant behaviors. Such dialogue allows for self-expression and respectful consideration of various perspectives that stretch beyond an individual’s boundary of knowledge and acceptance (i.e., paradigm) and toward a more holistic view of plants. It also offers new opportunities for creative, complementary approaches and collaborations to be conceived. There exist various methodologies to observe whether (and how) people can change their attitudes (and therefore beliefs) through strategic ways of communication [90]. Furthermore, a future systematic in-depth compilation of candidate plant intelligent behaviors could become a reliable basis for future scientific inquiry.

The strength of our results depends on the quality of responses and attentiveness of respondents. The time it took participants to finish the questionnaire ranged widely, from 20 min to 2 h. Repetitive responses may have partially reflected respondents’ waning interest, or point to structural issues with the questionnaire. The fourth research question could not be answered conclusively through our questionnaire. There is room for further development of our psychometric scale, which was a pilot design for a new study topic. Psychological metrics worth considering in future studies are those that address in detail (1) multiple relationships between people and plants; (2) intrinsic and instrumental valuations of other organisms; (3) the degree of understanding ecological interconnectedness; and (4) the level of trust and credibility in both scientific and traditional knowledge, which would potentially provide further insight into what types of ontologies shape researchers’ opinions of plant intelligence. Finally, our study was unable to test perspectives of plant intelligence as a general function (or model) of interactions between the three key components (and factors that contributed to each of them) of our conceptual framework (Fig. 1). We were limited to testing two variables (or groups) at each time independently. Future studies can elaborate on this. Regarding definitions of terms, our questionnaire did not clearly define what “using” TK systems in one’s field of study meant. As a result, we were unable to ascertain how respondents “used” TK in the way the study intended, as a respectful integration of different forms of knowledge.

This study demonstrates that the science, language, and subjectivity of participants all take part in forming and formulating their opinion of plant intelligence; therefore, these should not be considered or treated independently. Through the conceptual framework presented in this study, we hope to demonstrate that the understanding of plant intelligence extends beyond its dependence on scientific definition and knowledge. It is challenging to separate personal worldviews and thus subjectivity from conceptualizations and interpretations of plant behaviors in an academic environment. In addition, our results show some openness and support for intelligent behavior in plants while also, in general, showing a dependence on the conceptual framework and definition of intelligence. In conclusion, effective communication and an interdisciplinary dialogue among a wide range of interested scientists are actions that can help to close the divide and promote mutual understanding and consensus on plant intelligent behavior. Increased openness for and respectful considerations of TK can be a positive influence for plant scientists when interpreting plant behaviors in favor of intelligence. With growing respect and acknowledgement of diverse perspectives of plant intelligence, new research questions and approaches can be opened for future scientific investigation.