Even if laboratory experiments yield no conclusive evidence of mental states in worms, field observations may tell a different story. One area that merits further investigation is the process whereby earthworms construct their burrows. Burrow construction is an example of what Laland, Oddling-Smee and Feldman (1999) refer to as "niche construction" - the process whereby "[o]rganisms modify local resource distributions, influencing ... their ecosystems." (Other writers use the term "ecosystem engineering".) The term is a broad one: the authors describe plants as engaging in niche construction when they alter the temperature, humidity, fertility, acidity and salinity of their soils (1999, p. 10242). However, they recognise that some animals go beyond this behaviour when they construct artifacts such as nests, webs and burrows.
Charles Darwin spent forty years investigating the behaviour of earthworms, including their burrowing. Shortly before his death, he published a book (Darwin, 1881/1985) in which he argued for the existence of mental states in earthworms. Crist (2002) discusses his research. Darwin was well aware of the prevalence of instinctive behaviour, even in "higher" animals, as well as instances of what Crist (2002, p. 5) describes as "insect stupidity", which made him cautious about attributing mental states to invertebrates. Although he had not excluded the possibility of mental states in earthworms, Darwin was shocked by what he discovered from a study of their behaviour. As he wrote, "Some degree of intelligence appears a result which has surprised me more than anyone else in regard to worms"(1881/1985, p. 35).
A central feature of Darwin's argument is his employment of a "phenomenology of mind" (Crist, p. 7), in which he endeavoured to present the world from a worm's "point of view", and used this information to explain their behaviour, when more conventional explanations, such as instinct, proved to be fruitless. In cases where instinct was unable to explain an animal's behaviour, Darwin argued that it could be legitimately explained by analogy to that a human subject with similar sensory modalities (i.e. a similar "point of view"). If we would be prepared to credit a human subject, who had a "point of view" like that of the animal and who behaved like the animal in similar circumstances, with intelligence, then we should credit the animal with intelligence, too.
Darwin was struck by the fact that earthworms plug the openings of their burrows with leaves and petioles (leaf stems). He performed experiments to see how earthworms handle leaves. While he acknowledged that plugging holes was instinctive in earthworms, he argued that they displayed a primitive intelligence in the way they handled leaves (including leaves foreign to their habitat, such as pine leaves, and leaves artificially constructed from paper by Darwin). He concluded that they were able to gauge the general shape of a leaf by touching it with front end of their bodies, and that they used this information to manipulate leaves more effectively. He employed his phenomenological method to argue that the behaviour of worms resembled that of people with a highly developed sense of touch:
Darwin's findings are presented in the Appendix.
Assume that the research Darwin conducted over a period of forty years was accurate, the questions we have to ask are: (1) Is Darwin's methodology appropriate for identifying mental states? and (2) Do Darwin's findings establish the occurrence of mental states in earthworms?
It would be easy to ridicule Darwin's "phenomenology of mind" (as Crist calls it) as anthropomorphic. In my opinion, such a criticism misses the point. To some extent, humans and other animals share what Wittgenstein called a form of life. This, surely, is the starting-point for any investigation of animal behaviour. Darwin began his monograph on worms with a discussion of what he called their habits - i.e. their habitat, sensory capacities, food and digestive systems. Before we can understand the mind of any creature, we must first understand how it lives and what it habitually does. Of course, it would be illegitimate to infer that similarities in human and animal behaviour require similar causes, but that is not what Darwin did. He identified patterns of behaviour in earthworms, that in humans would be ascribed to intelligence, and then proceeded to look for alternative explanations. Only when he concluded that these explanations were deficient did he tentatively propose that worms possessed a rudimentary form of intelligence.
However, two cautionary remarks are in order. First, Darwin's use of analogy with human cases (e.g. people born blind and deaf) works only for those senses that are shared by humans and animals - e.g. touch in the above case. It could not be used for animal senses that humans lacked - e.g. sonar in a bat.
Second, similarity in behaviour is a matter of degree. A man born blind and deaf (to use Darwin's example) may rely heavily on his sense of touch, as an earthworm does, but he is much larger than an earthworm, and does not live below the ground. The use of inter-subjective analogical reasoning becomes problematic when there are significant differences in the subjects' habitats.
Notwithstanding these criticisms, Darwin's argument for mental states in earthworms remains plausible. There is, however, a valid criticism that can be made of Darwin's methodology: in his quest to identify mental states, he posed a false dichotomy between intelligence and instinct (which he conceived as an invariant pattern of behaviour), and assumed that any behaviour which could not be ascribed to instinct or chance must indicate a rudimentary form of intelligence. This becomes apparent in his reasoning about earthworm burrows:
Darwin assumed that the hallmark of intelligence is flexible behaviour, and that instincts are utterly invariant. The conclusions reached in this chapter regarding flexible behaviour patterns expose the gaps in Darwin's reasoning here. As we saw in Conclusion N.11, "inflexible" is not the same as "invariant". A fixed action pattern does not imply fixed behaviour. Adaptive behaviour in unforeseen circumstances may seem to be a sure sign of intelligence, but it does not necessarily indicate true flexibility, as we saw in our discussion of bacteria. A fixed rule may generate behaviour which fortuitously turns out to be adaptive and can even be applied to novel, non-native species like pine leaves - e.g. "drag a leaf by the base if and only if it has a narrow base". Could earthworms be programmed to follow such a rule when plugging their burrows?
Finally, even if earthworms are shown to exhibit flexible behaviour patterns, this does not necessarily show that they have cognitive mental states. Internally generated flexible behaviour is a necessary condition for cognitive mental states (Conclusion N.12), but that does not make it a sufficient one.
What I would suggest is that Darwin asked the wrong question about the behaviour he observed. With the benefit of sharper philosophical terminology, we can formulate one: what kind of intentional stance is required to explain this kind of behaviour? Will a goal-centred intentional stance suffice, or is it more appropriate to view earthworms as agents?
A useful way of answering this question might be to ask whether the rules earthworms follow are learned or not. For instance, do immature earthworms drag leaves to their burrows in the same fashion as mature specimens, or is there a "learning curve", where they learn through trial and error what method of dragging works and what does not? As far as I can tell, neither Darwin nor subsequent researchers have endeavoured to answer this question. This would be a fruitful area to investigate. A "learning curve" with regard to rules would suggest that earthworms are agents, which are trying to plug their burrows with a minimum of effort. Such an explanation would be amenable to an agent-centred intentional stance. If, on the other hand, no learning takes place, there seems to be no good reason to explain their leaf-pulling behaviour in mentalistic terms, and a goal-centred intentional account of this behaviour should be adequate.
It may be well to remember how perfect the sense of touch becomes in a man when born blind and deaf, as are worms. If worms have the power of acquiring some notion, however rude, of the shape of an object and of their burrows, as seems to be the case, they deserve to be called intelligent; for they can act in a manner as would a man under similar conditions (1881/1985, p. 97).
If ... worms acted solely through instinct or an unvarying inherited impulse, they would draw all kinds of leaves into their burrows in the same manner. If they have no such definite instinct, we might expect that chance would determine whether the tip, base or middle was seized. If both these alternatives are excluded, intelligence alone is left; unless the worm in each case first tries many different methods, and follows that alone which proves possible or the most easy; but to act in this manner and to try different methods makes a near approach to intelligence (1881/1985, p. 65, italics mine).
*** SUMMARY of Conclusions reached