Minds on Monday: Seeing and Doing

What we see as a species depends on what we do as a species. This was the teaching of the visionary (though often a little too much so) Estonian physiologist Jakob von Uexküll (1864-1944). Uexküll posited that the world sensorily revealed to an organism, its Umwelt as he called it, consists of environmental features constructed (or selected, as I would prefer to say) for their relevance to the organism’s behaviour and survival.

A honey bee, for example, divides flowers into “broken forms” such as crosses and stars and “compact forms” such as circles and squares. This is significant because bees collect pollen from open flowers, which have broken forms, not from buds, which have compact forms. I think Uexküll meant to say that these two forms are directly perceived—i.e., not simply indicated by a more complex featural substrate. Looking at the objects pictured above, the bee visually experiences just a simple differentiating quale, which I represent by X and O. (It’s actually quite implausible that bees see just a differentiating quale: after all, flowers have evolved special colour markings to attract bees. But perhaps Uexküll wasn't much concerned about direct perception and qualia.)

Perception is species-specific. Take colour vision. It is now generally believed that this perceptual capacity in primates co-evolved with the colour of fruits and berries. Monkeys live on fruit; fruit-bearing plants depend on monkeys for the dispersal of their seed. It is therefore advantageous for monkeys to be able to find fruit, and advantageous for plants (of a certain sort) to have their fruit be findable by monkeys. So fruit evolved toward red or yellow colours to stand out against green foliage. Monkeys co-evolved an extra cone-cell that enables red and yellow to be sharply differentiated from green. When monkeys scan a scene, red fruit perceptually “pop out” from the background of greenery, thus reducing search time (particularly where there is a high spatial frequency of colour variation). Colour vision evolved for a very special purpose, and its characteristics are correspondingly special purpose. It is good in "dappled and brindled" scenes, as John Mollon (an originator of the co-evolution hypothesis puts it.)

Primate colour perception is quite specific: its space of similarities is shaped by the use to which it is put. And it’s not just that primates are more sensitive to some colours than to others. It’s also that the similarities and differences between colours evolve in a way that is dictated by their way of life in a specific environment. Birds, which evolved colour vision independently of primates—they have no common ancestor with colour vision—are different again. They use colour not only in the search for food, but also for mate selection and navigation. Since short wavelengths are scattered more than long, the sky is bluer in directions perpendicular to the sun. (Light in that direction is scattered light.) Inhabiting, as they do, a three-dimensional environment without landmarks, birds use shades of blue for directional orientation. Then again, many fish use yellow filters to enhance blue-green contrast. It is not colour that we need to perceive as such: rather, certain colour contrasts in objects of practical interest.

Nor is perceptual specialization restricted to colour or to secondary qualities.

• Conspecific “voice” recognition is important in communicative species, among whom humans mark an extreme. Our ability to make speech sounds (which, by the way, comes at the cost of a gullet more prone to choking than that of apes) would be useless if not matched by our ability to decipher these sounds. And vice versa. We can’t easily decipher bird song; birds cannot differentiate our vowels and consonants.
• Humans are bad at estimating vertical dimensions, but better at horizontal size and distance along the ground. It’s dubious that birds suffer from this distortion. This, of course, has to do with our respective means of locomotion.

None of this should obscure the range of activities that animals undertake. Unfortunately, though, it often does. Uexküll and Gibson focussed on the sensory guidance of the body. Proud of their departure from conventional wisdom, they forgot the defining contribution of the senses—to learning and deliberation. The Uexküllian Umwelt and Gibsonian affordances are environmental features directly relevant to bodily actions—a bird sees a branch of a tree or an electrical cable as affording it a perch and a cornfield as affording it nourishment. But the defining characteristic of a sense modality is that it contributes to consciousness. So, for example, vestibular information concerning bodily position is usually not counted as sensory because it operates largely outside of consciousness and does not contribute to learning or deliberation.

Vestibular and proprioceptive information operate outside of consciousness and outside of the organism’s choice. Suppose you are hungry and see a restaurant on top of a hill. You may choose to climb the hill, or you may decide it’s not worth the effort. If you do climb the hill, your body posture and movements will adjust to the terrain without your explicit choice or calculation. It is characteristic of senses such as vision and audition that they act in the former way: they present information that is held for unforced and optional decisions (which is not to deny that they also contribute to some that directly control the body.

You see a red apple. Gibson says you see an affordance: it’s eatable. But in its redness you also see an indication of ripeness. Will you infer it is ripe? That depends on other indications: perhaps you think it only looks red, or that it has been painted red. Either way, the sensed redness is an epistemic affordance: it allows you to learn from experience and to apply previously acquired knowledge.