How to teach a blind extraterrestrial to see

by Dave Maier

Dark OrbitCarolyn Ives Gilman’s 2015 novel Dark Orbit is intensely concerned with the extent to which reality outstrips or transcends our knowledge and/or sense perception. Indeed, the jacket copy tells us that one character’s “most difficult task may lie in persuading the crew that some powers lay [s/b “lie”?] beyond the boundaries of science.” I generally don’t like this sort of talk, which smacks of obscurantism. Naturally if you construe “science” narrowly enough – identifying it with this or that set of procedural or substantive commitments – some aspect of reality will probably remain opaque to “science” so construed; but that doesn’t tell us much.

While some of its characters haven’t quite thought all this through, Dark Orbit itself is quite thought-provoking and well worth reading for this and other more conventional sci-fi reasons (like that it’s a cracking read). I found particularly interesting the conceptual difficulties the characters run into when dealing with the nature of the senses and their relation to the reality beyond our heads. Naturally this discussion may involve some **SPOILERS**, but not, I claim, anything particularly serious.

* * *

I don’t want to get into too much of the plot of Dark Orbit, but by p. 128 the crew of the Escher is orbiting a distant planet, which they name Iris, and run into some curious flora.

[Botanist Hua Ming] showed a picture of a plant like a globe on a stalk, crowned by a ring of sharp spikes. “This is the type specimen of the first new species discovered on Iris,” he announced. “We have named it a brickle, species name Brickellia.” […] He then showed a slide of a seemingly identical plant. “This is Impedomia. The difference, you see, is that there are five spikes instead of seven.”
Everyone was silent. Hua Ming was about to continue when Sara spoke up. “What’s the significance of having five spikes?”
Ming stared at her for a second, then explained, “It’s the difference between two otherwise identical plants.”
“But why is it so meaningful?” Sara pressed.
“Because it puts them into different species.”
“But you just made up those species.”
“No, I didn’t. The plants are different, can’t you see?”
“Couldn’t there just be five-spiked brickles and seven-spiked brickles?”
“No, because the five-spiked ones are impedomia.”
Sara gave up, and Hua Ming continued on with his pictures of brickle meadows. Sara found herself mentally sorting out the species in his pictures. It was satisfying, because the comprehension-defying mix of Irisian plants was now composed of known things, brickles and impedomia. Known things about which nothing was known but their names—still, it gave her a sense of mastery.

Seen in the context of Dark Orbit’s presumed theme of cutting science down to size, this bit seems like a shot at our – or at least science’s – presumed tendency to mistake our own invented categories for objective features of reality. Surely reality cannot, as Ming suggests (“the plants are different, can’t you see?”), determine by itself whether it is right to say that we have one species here or two, especially since the only difference he mentions is the number of spikes on each. Yet Sara finds Ming’s categorization “satisfying” in giving her “a sense of mastery” where it still seems that she has no real knowledge of how things are over and above what we’ve decided to say.

Ming can do better. He cannot simply point to the number of spikes, but must also explain (as Sara notes) why we should regard that difference as one between two species, as that term is properly used. This is a pragmatic consideration rather than a metaphysical one, so it may still seem that we are imposing our own needs on the world rather than grasping its objective nature; but in so doing we are bringing to bear on the issue our knowledge of nature as well. Sara’s reaction at the end of the scene helps to explain why we may feel uncomfortable leaving the issue open until more data come in; still, the discomfort seems misplaced, as well as uncharacteristic of actual science.

This issue of the apparent arbitrariness of categorization comes up again a bit later. Iris is home not simply to brickles and/or impedomia but also native humans, who conveniently speak a version of “Universal”. One native, Moth, is taken aboard and immediately discovered to be blind. As we know, however, while Escher’s crew does not, Moth’s people have lived in a pitch-black cave from time immemorial, and consequently “see” things rather differently (thus raising another cluster of philosophical issues we won't have time for today). Unlike most Irisians, though, Moth occasionally goes outside; but while it is light there, and her eyes happen to work fine, she ignores the meaningless and distracting “blotches” before her eyes, and indeed must often close her eyes and listen carefully lest she get lost. Since Moth believes that “sight”, as described by Sara, is very valuable, the scientists try to train her to attend to the “blotches” instead of ignoring them, in order to learn to see. (Yet another concern of Dark Orbit is whether and how it may be unnecessary or counterproductive or unethical to train the Irisians to see in this way.)

Not surprisingly, this process turns out to be fairly difficult. As we join her here, Moth is having trouble recognizing the same object from different angles.

Color was the first thing to cause confusion. Moth insisted that the blue cup was a different color when viewed from the chair than it was from the door. Sara told her, “No, it’s the same color, it’s just the lighting that’s different.”
“So I must memorize many colors thou dost call blue?” Moth said in frustration.
Sara remembered her own frustration with Hua Ming’s arbitrary botanical categories. Yet she had accepted the category of “blue” not as if it were imposed by humans, but as natural, self-evident.

Here, it seems to me, our characters are conflating a number of rather distinct philosophical issues, and/or making certain questionable philosophical assumptions. Let us unpack.

First, Sara seems to believe that our categories are either “natural” (i.e. accurate to reality) or “imposed” (i.e. arbitrary and thus presumably most likely inaccurate). But categories, like language itself, simply make claims possible; they aren’t claims themselves. In employing our concepts, we generally tweak them as necessary, the better both to manifest our beliefs about the world and to communicate them to others for a particular purpose. Naturally our interlocutors may dispute our claims; but in so doing they need not dispute as essentially subjective (“imposed”) the concepts we use to make them. Nor are they simply “imposed” by nature either, as (just like claims themselves) they are generally revisable.

Another issue is that while Sara mentions Ming’s botanical categories, the actual context is one of color terms. In other words, she is ignoring the traditional distinction between “primary” and “secondary” qualities of an object, where the latter make essential reference to how things appear to us, while the former reflect how things are independent of our senses. Traditionally, “secondary” qualities are subjective and thus in some sense not real, while “primary” qualities are the objective and thus metaphysically legitimate ones. While not entirely spurious – in certain contexts it is virtually unavoidable – this distinction, I would claim, is not obligatory and certainly does not have the metaphysical and epistemological significance claimed for it.

So when Sara conflates these things, this doesn’t seem to be the problem with her response. Instead, I’d start by getting Moth to distinguish between a) philosophical or theoretical questions about sense perception in general (which would thus apply to senses she is familiar with, like hearing and touch), and b) questions about sight in particular, which concern how she’s supposed to deal with the blotches in front of her eyes.

This means that the problem of determining the color of the same object in different lighting is not the same as that of the seeming arbitrariness of color categories. With respect to the former, Moth seems to be thinking that she must first identify the particular shade of the appearance and then remember which ones correspond to our color term “blue”. She does not realize (and Sara does not explain) that with practice, she will learn to (as Helmholtz puts it) “discount the illuminant” and actually see the varyingly lit object as the same color throughout. This is true no matter how many different color terms we use.

But in this sense also, Moth needs to relax. It doesn’t matter how many “different colors” she needs to learn, although she’s right that it will take some time to get good at it. When children learn to use color words, they tend to call lots of different things the same color (often “brown”, as I recall) at first, and then settle into the main seven or so. After that it’s a trade-off: we can perfectly well lump a whole bunch of shades together as “green”, even if Sherwin-Williams employees call them “forest green” and “lime green” and “Kelly green”. What you say depends on whether you’re telling me which tie to wear (“the green one”) or painting your house – and even in this case, we don’t need to worry, that’s why they have swatches. Again, this doesn’t have anything to do with figuring out what colors things are in changing light.

Alas, more confusion is in store for poor Moth.

Moth also had trouble telling which boundaries were the significant ones that defined the edges of objects. […] On the second day of lessons, Sara and Moth got into an argument about why the legs of a table, which were separated from the top by a very distinct boundary, ought to be classed as part of the table, instead of something different.
“Because they’re attached!” Sara said in frustration.
“I cannot tell that,” Moth said.
“Not from where you’re standing. But if I move the table, the legs will come along. You know that.”
She looked thoughtful. “So the table-symbol in the eyes is supposed to represent things about the real table?”
“No, the table in the eyes is the real table.”
“How can that be? Doth the table send out thoughts about its nature that the eyes receive?”
That question led to a long explanation of electromagnetic radiation. At the end, Moth said, “So what the eyes see is the radiation bouncing off the furrniture?”
“That’s right.”
“Not the furniture itself. So how can thou be sure the radiation is truthful?”
“Experience, Moth. Just trust me on this, okay?”
But when Sara reported this conversation to David, he said, “You know, she has a point. Sight is a secondhand sort of sense. We can’t be sure something isn’t lost in translation.”
Sara groaned. “Don’t you get all epistemological on me, David!”

Sara’s first point is a good one. Moth knows what a table is already (i.e. by touch), and that if she picks it up by the top, the legs will follow. The point doesn’t even depend on the legs being attached: even if the top is completely detachable, the legs are “part of the table”, and Moth should know this conceptual truth already. In some contexts, of course, we will regard the legs as distinct from the top – that’s why we have a name for them, after all – but not from the table itself. Sara should brush aside Moth’s quibbles about sight here and instead try to make sure they agree about what “tables” are.

The next bit, which Moth seems to think follows from the first, is another thing entirely. That “the table-symbol in the eyes” – let’s call this a bunch of “sense-data” – is (merely) a “representation” of the “real” table out there in the world is a commonplace of modern Western philosophy. Again, though, it would be better if Moth approached this conceptual issue via a sensory modality with which she is already familiar. Touch can be confusing in this respect, as since we need to be in physical contact with an object in order to feel it, we are less likely to think of the sensation of touch to be in any significant sense removed from the object itself (even though, as a sense-datum, it is just as much “in here”, conceptually speaking, as a sensation of vision).

But Moth hears sounds too, and presumably gets that when one hears (the sounds of) someone pouring a glass of water, we can perfectly well say, depending on what exactly our point is in the context, that we “hear” either thing: “I hear someone pouring water” or “I hear the sound of water being poured”. By analogy, then, Moth shouldn’t balk at saying, when she comes to understand the blotches for their actual significance, that she sees not just the light bouncing off of the table, but the table itself. (Sara could be clearer here.)

But is it the “real” table? Here we return to the philosophical issue, common to all sensory modalities, of the epistemological status of sense-data generally (as David points out), as well as the metaphysical status that we often take to go along with it. We’ve already isolated our real issue: it doesn’t have anything to do with color, and it doesn’t have anything to do with sight. Given the “subjectivity” of sensory experience, how can we take it to give us knowledge of an objective world?

Sara naturally appeals to our experience in reply. While our senses do indeed deceive us occasionally, we eventually learn when this is most likely (e.g. mirages in the desert) and when it is merely a conceptual possibility not worth the worry. In fact, when we come to understand mirages, we generally think of the phenomenon not (simply) as a false impression of the presence of water (“I see what looks like water, but I know there isn’t any there”) but instead a veridical impression of a mirage (“I see a mirage”). Again, once we know how to use the relevant terms – even in this apparently quintessential example of sensory malfeasance – there’s no reason to say that there is in fact “no mirage there,” even if you have to be standing in a certain place to see it (compare things like rainbows or “magic eye” optical “illusions”).

But Sara’s answer is unnecessarily misleading. She makes it sound like experience gets us across an acknowledged epistemic gap. This leaves an opening for all sorts of Descartes-inspired skepticism: for it is the trustworthiness of experience that we were worried about in the first place. We shouldn't be the guy who (borrowing Wittgenstein’s words from another context), because isn’t sure he can believe what the newspaper says, buys another copy.

The better answer is to deny or at least reconstrue the gap in question. The senses are causal intermediaries between mind and world – our knowledge about the world depends on them – but this does not make them epistemic intermediaries. This latter, broadly Cartesian idea fails to explain why the thoughts we take to be beliefs standing in need of epistemic justification actually have content at all. For where would that content come from if not from the world and its constituent objects, to which our thoughts thereby refer? That we have language at all, and thereby have contentful beliefs to be true or false, makes broadly-Cartesian radical doubt unintelligible.

Given what is thereby revealed as an epistemically unmediated relation between mind and world, what actually needs explaining here is not the conceptual possibility of knowledge of the real, given the “unreliability” of the senses, but instead the conceptual possibility of error. Why, for example, if the meaning of our words depends on referential connections to the world, is the belief that unicorns exist not simply intelligible but also false?

Naturally we cannot burden poor Moth with all this just yet; but once she learns to read, we'll sit her down with the complete works of the ancient Terran philosopher Donald Davidson, and she'll be all set.

Like what you're reading? Don't keep it to yourself!
Share on Facebook
Tweet about this on Twitter
Share on Reddit
Share on LinkedIn
Email this to someone