Not only vertebrates colonised the land in the early Palaeozoic, but, millions of years earlier, arthropods (insects, spiders, etc.) also made the sea-to-land transition (Lozano-Fernandez et al, Reference Lozano-Fernandez, Carton, Tanner, Puttick, Blaxter, Vinther, Olesen, Giribet, Edgecombe and Pisani2016). Some also developed excellent eyesight. Was it possible that they, too, acquired consciousness? One cannot, however, assume that just because different groups of organisms face similar evolutionary pressures, they will adapt in a similar fashion. Is there evidence that any Arthropoda plan actions based on representations of the world in other than its current state?

Jumping spiders (Salticidae) are hunters and rely on vision at a distance. Species of jumping spiders, Portia, prey on other spiders and display especially sophisticated hunting tactics. A single species often hunts multiple types of prey – web-building spiders, other Salticidae, and insects. They use different tactics for each type of prey, stalking, making detours to wait and ambush prey, manipulating the webs of web-building spiders to suggest a catch is present, and using different tactics for the males and females of some species (Jackson & Pollard, Reference Jackson and Pollard1996). Selection of appropriate tactics relies upon identifying the type of prey from a distance. Electrophysiology (Menda et al, Reference Menda, Shamble, Nitzany, Golden and Hoy2014) reveals neurones in Portia selective for species-specific visual features. Cues to prey-type result in endogenous selective attention for such prey (Cross & Jackson, Reference Cross and Jackson2009). Portia’s flexibility in hunting, stalking, and ambush has led them to be described as “eight-legged cats” (Harland & Jackson, Reference Harland and Jackson2000)!

Cross (Reference Cross2024) describes the behaviour of an individual Portia in the wild. The spider, on seeing its prey, a web-building spider, did not approach directly. Instead, it took a long detour, eventually appearing directly above its prey, which it then dropped onto and devoured. This tactic prevented the prey from seeing its adversary or from feeling it approach across its web and seems like planned behaviour. The best evidence for planning and maintaining a representation of the world other than that presently observed, though, comes from laboratory experiments.

Fleming and Michel (F&M) note the evolution of flexible necks after tetrapods transitioned to land. Portia do not have flexible necks but have an “eye-movement” system much like ours – six muscles allow gaze to be directed over 60 degrees of visual angle by moving the retinae rather than entire eyes. These retinae possess fovea, so acuity in central vision is far greater than in the periphery, and hence, central vision is selective. Portia’s gaze movements are like ours, making both saccades and smooth pursuits. They also make systematic scanning movements to cover the visual field (Land, Reference Land1969; Hill, Reference Hill2022).

Tarsitano & Andrew (Reference Tarsitano and Andrew1999) exploited this in a study where spiders’ gaze was tracked with video. From an initial position, spiders could take one of two pathways, only one leading to a reward. In some trials, there was a gap in one pathway. When a gap was present, the spider initially spent more time looking at the pathway with the gap, but subsequently spent three times as much time looking at the intact pathway. Nearly all the spiders tested then took the intact pathway to the reward. This certainly suggests planning based on initial observation of the gap.

Tarsitano & Jackson (Reference Tarsitano and Jackson1997) and Cross & Jackson (Reference Cross and Jackson2016) conducted studies in which Portia initially viewed a test apparatus from a tower. From this vantage, they could see the layout of two pathways leading to containers, only one of which contained a reward. The reward, a dead prey spider, had been soaked in alcohol, then dried, and sealed under a layer of adhesive. It provided no odour cues to its location. The pathways were complex, with turns in horizontal and vertical directions. Once the spider climbed down from the tower, it could no longer see the contents of the containers. There were multiple variants of the apparatus with pathways of different lengths, sometimes crossing over one another, and starting at different positions relative to the tower. One pathway led to reward in half the trials, the other in the rest of them. Spiders received no training and were only tested once. In all but one of the arrangements, spiders were significantly (p<0.001 for each arrangement) more likely to select the pathway leading to reward (spiders were reluctant to start on either pathway in the one exception). To achieve this, the spiders must have determined which pathway led to reward and where it started on the ground, all while they were in the tower. They must then have retained this information and translated it into behaviour leading to success at ground level – using a representation of the world seen from one viewpoint in planning action when the current view of the world is from a different viewpoint. Exactly the circumstances that F&M suggest demand reality-monitoring, and hence, consciousness – in a poppyseed-sized brain with just 100,000 neurones!

This evidence lends credence to the suggestion that some arthropods might be conscious (Chittka & Wilson, Reference Chittka and Wilson2019), but does F&M’s thesis imply that this is phenomenal consciousness? Might representations be entirely propositional, for example, “there was a reward on the left,” but tagged with some indicator that this does not reflect the current state of the world? How can we distinguish between a propositional representation of the state of the world and a phenomenal one? One possibility is to test organisms’ responses in situations where propositional and phenomenal representations are in conflict. Consider, for example, a London Bus moving from shadow into bright sunlight. The phenomenal experience of the colour elicited by the bus, its colour appearance, changes – it becomes less blueish and brighter. But our belief that it is an object painted in standard London Bus red (Pantone 485C) remains constant. If we could train organisms to respond either to phenomenal changes or to changes in objects’ surfaces, and show that these can dissociate, then it would be hard to conclude anything other than that the organism was phenomenally conscious. Alas, I doubt Portia, smart as they seem, could learn such tasks. Maybe their “consciousness” just is not like ours?