Plant motor intentions challenge definition of cognition

Plants can communicate, remember, recognise kin, decide, and even count — "all abilities that one would normally call cognitive if they were observed in animals." 

These were the conclusions reached by Umberto Castiello of the University of Padua in a 2020 review of recent experiments on plant capabilities. Now, Castiello is a co-author of new work in the Journal of Comparative Psychology that reveals evidence of different 'movement intentions' in plants, depending on whether they are growing alone or in the presence of another plant. 

The team writes that the new findings "support the idea that plants are capable of distinct social intentions and take action accordingly, exactly as humans do." Though this may sound strange, plant cognition is a growing field of research, and one that poses a direct challenge to the traditional definition of cognition. These new results feed in to the ongoing controversy. 

The new work was led by Bianca Bonato, also at the University of Padua. It focuses on snow peas — climbing plants which need structural support, such as a stick, to survive and thrive. To search for a support, the snow pea puts out tendrils, which move in a helical exploration pattern. Once a tendril detects a suitable support, it grasps it. 

This experiment had two conditions. For the 'individual' condition, the researchers planted 12 snow pea seedlings in individual pots, with a support stick initially positioned 12cm away from the seedling's first leaf. For the 'social' condition, they planted plots with double the soil volume with two seedlings. A stick was positioned mid-way between each pair of plants, again 12cm from the seedlings' first leaves. The team then made video recordings of the plants, to track their growth. 

The researchers observed clear differences in the 'reach-to-grasp' tendril movements between the two conditions. For snow peas that were growing alone versus in pairs, the tendril movements were faster, and the amount of time spent in helical explorations was shorter. This suggests that the plants growing in pairs were more "cautious" in their movements, the researchers write. Research on humans also finds more careful reach-to-grasp movements in a social context. 

For each pair, there was a 'winner', which managed to securely grasp the support, and a 'loser'. The winners' tendrils moved faster and remained closer to the support for longer, before entering the attach stage. The actions of the 'loser' plants were, in contrast, slower and less directed towards the support — more 'submissive', the team argues. 

Why one plant should become dominant in the endeavour to secure a support, while the other becomes submissive, is not clear. An analysis of the plants' roots showed that both members of a pair had a similar root area, and so similar access to water and nutrients. This suggests that the two plants were not actively competing with each other. "Rather, the two individuals share resources and act in concert for the apparent benefit of one of them," the team writes. 

They also observed that when one member of a pair successfully grasped the stick, the other started to direct its tendril explorations in another direction, investing its energy in a new search for a support. 

These new findings expand the list of plant abilities that would be called cognitive if observed in animals. When animals intend to do something physical, they start to move their muscles. When plants 'intend' to do something — such as move towards light or heat, as well as to grasp a support — they do the plant equivalent, generating action potentials and synthesising proteins for the required growth. 

This new work suggests that, as happens with animals, plants modify their intentions and their behaviour according to the social context. For Bonato and her colleagues, the fact that the plants' actions — their growth patterns — were different in social versus individual conditions is evidence of different underlying intentions, although the precise biological mechanisms driving these intentions are still unknown. Their behaviour is, the team argues, deliberate and purposeful.

Not everyone will agree with the team's analysis, of course. According to the classical conception of cognition in animals, acting in a goal-directed way requires first perceiving the environment, which plants, with their many senses, certainly can do. Then, neurophysiological mechanisms in the brain are used to create mental representations — including concepts and intentions — which guide physical actions. Clearly, plants don't have brains. For this reason, many find it difficult to entertain the idea that plants could show evidence of cognition. 

However, others argue that cognitive activities don't depend on mental representations. According to an argument outlined in a recent paper by two philosophers, active exploration of the environment is evidence of cognition. Bonato and her colleagues argue that if we set aside the requirement for a representation of the environment in the brain, this new work shows that plants can act in an intentional, goal-directed manner — and that they can be considered cognitive agents. 

Recent studies have also shown that some animals that were once thought to be simple automata are capable of complex learning and planning. This has been found for spiders, for example, which have only miniscule brains. For some researchers, this new work will push plants, too, further along the path to cognitive recognition. 

The new findings open up all kinds of questions that the researchers acknowledge would have sounded absurd only a few years ago. These include: to what extent exactly can a plant plan its movements? And, are the basic processes involved in the intention to interact socially the same for plants as for animals?

Future research will have to explore these questions. For now, the concept of plant 'intentions' will remain controversial. For some, though, the new work will be seen as an exciting step in the broader evolution in our understanding of plant functioning.

Read the paper in full: https://doi.org/10.1037/com0000351