Tuesday, February 10, 2009

On Attributing Consciousness

Alderson Warm-Fork of Directionless Bones has a post up about how people attribute consciousness, which is a subject I've been thinking about a lot recently, so I'll just go ahead and flesh out the responses I started to leave in the comments section there.

Alderson begins by pointing out that much of Western philosophy has proceeded on the assumption that everything not "obviously conscious" is inanimate. The reigning worldview being a purely mechanistic one*, philosophers didn't spend a whole lot of time pondering how other minds (indeed, minds in general) worked until later.

Now I want to suggest a different model, where attributing consciousness is central. The model is something like this:

1) When we perceive movement, we by default attribute consciousness (or ‘life’, or some such vague idea of being something more than just a rock).

2) When we perceive that a certain movement can be entirely comprehended by us, we retract that attribution of consciousness.
In other words, rather than starting from an assumption of insentience, we ascribe sentience provisionally to an entity based on a very inclusive set of criteria (e.g., it appears to move under its own power), with the understanding that we might revise this opinion later should that entity fail a more rigorous test.

I think this is true enough, although I have some reservations about the phrasing (both my own and Alderson's). Laying it out logically gives the reader the impression that these are conscious, intellectual processes, which they usually aren't. Usually, the whole tree gets resolved too quickly even to make it into a form we would recognize as "thought"; it's a series of choices between patterns of reactions, and the way we typically react to things we believe to be conscious (i.e., living things capable of volitional movement) is to focus our attention on them and try to divine their intentions (or, if "intentions" is too strong a word, their likely course of action, e.g. Is that squirrel going to run out in front of my car?).

Commenter Awais did not think this theory quite covered it, however.
[T]he mechanism [by which] I believe people attribute consciousness to others is that of logical analogy with their own behavior: we attribute consciousness to objects to the degree to which they approximate human behavior.
I think there's something to this one, as well, but I think it's another set of branches on the tree Alderson describes, rather than a replacement for it. There's a lot of ambiguity in the words we use to describe mental processes, and that confusion definitely appears in both the original post and the comments. Alderson used "consciousness" interchangeably with "life" (as in, living/conscious things move, while nonliving/unconscious things do not move unless acted on by an outside force) while Awais's usage of "consciousness" had more in common with "intelligence" or "sentience". I am using Alderson's terminology, more or less, although I have a few synonyms thrown in to confuse things even more. (I use "sentience" and "consciousness" more or less interchangeably). Anyway, I think Awais's criterion, the approximation of human behavior, comes into play more when people decide whether to attribute intelligence, which goes beyond basic capacities like perception and volition to enable such complex behavior as language, social organization, adaptation and invention.

I think the idea that we look for "humanlike" behavior as evidence of intelligence needs some poking at, though. As I point out in my contribution to the thread, ideas of what is "human" vs. what is "animal" have a long history of being used to justify various systems of oppression on the grounds that only the oppressors were in possession of full humanity, and thus the oppressed groups needed their enlightened patronage to keep them from sinking into their natural, animalistic state of indolent squalor. Accordingly, the usual things that come into our heads when we try to come up with uniquely human qualities tend to reflect the values of wealthy, educated, propertied Western white men.

Aside from being bounded by culture, our conceptions of what is "human" are also limited by our own lived experiences. The philosopher Ian Hacking gives an illuminating example of this in his talk "How We Have Been Learning to Talk about Autism," when he mentions the limitations of that capacity for immediate, unintellectualized insight into another person's state of mind that we now call Theory of Mind:
[Gestalt psychologist Wolfgang] Kohler pointed to a wide range of phenomena in which we "see," but do not infer, what a person is doing. Among these are ... seeing [a] child wants to touch [a] dog but doesn't dare, seeing that [a] friend is startled by something, and seeing where that something must be, [and] seeing that a man is upset by an unwanted task. Let us call these "Kohler's phenomena." You can think of innumerable examples. These phenomena, so familiar to most people, are precisely what is not familiar, automatic, immediate or instinctive for most autistic people. Expert observers reports that autistic children don't see someone is in a bad humor; they don't follow the direction of someone's gaze; they do not readily understand what another person is doing; that is, they do not readily recognize intentions. Conversely, most people cannot see via the behavior of severely autistic people what they feel, want or are thinking. Even more disturbing is an inability to see what they're doing, what their intentions are. Their intentions make no sense. With the severely autistic, it may seem as though they don't even have many intentions. (boldface emphases mine, italics are meant to approximate his emphases)
In other words, there are limits on the degree to which people's thoughts, feelings and intentions are apparent from their behavior, and these limits are different for different people.

As I say on Alderson's site, this inability of NTs to "see" the reasons for autistic people's behavior usually leads them to conclude that there are no reasons for it:
We move, we act ... but our movements and acts have no recognizable goal, and thus people assume we lack intelligence, and lack all but the most rudimentary stages of consciousness. Our emotional responses are similarly discarded as meaningless, because we do not react in the same way most people do to the same things. Things in the environment that most people might not even notice scare us or irritate us, but because the stressors don't even make it onto most people's radar, we are assumed to be throwing a fit for no reason. So our movements, our behaviors, and even our emotional responses and attempts to communicate are discarded as meaningless and we are believed not to be conscious or intelligent to the same degree that most people are.

Consider the phrases most often used to describe autistic people: we're "in our own worlds," "cut off from other people/the world around us," "unresponsive," "noncommunicative," and so on. Even the things we do entirely for our own enjoyment fall under this semantic umbrella of meaninglessness and sterility: our interests are "perseverations" or "obsessions," and our forms of solitary play are "stereotyped," "repetitive" rituals. Like some of the earlier instances I mentioned of privileged groups of people defining other groups of people as less human than the privileged group, as lacking in some core component of full personhood, autistics in this framework are seen to lack something essential to human nature. But unlike those other examples, we are not likened to animals**. We are not considered subhuman so much as inhuman.

*Of course, most of Western philosophy, at least before the "Death of God" in the nineteenth century, has also been very strongly Christian-influenced, and accepted the existence of a spiritual dimension. I've only just now pondered how weird this is in light of the way Western thought has traditionally viewed the natural world --- i.e., as an inert substrate to be acted upon by conscious beings, i.e. humans.

**There is one way in which I think autistics and animals have been similarly treated with respect to attributions of consciousness: we have both been reduced to our behavior, and then been subjected to all manner of awful, demeaning, and/or surreal things in the name of shaping said behavior.


thinkingdifference said...

that's interesting. i also found myself thinking about religions and attribution of consciousness. the idea of smth moving (and until u understand the 'causes', you temporarily ask if it's conscious) - in some worldviews, it can legitimately be thought of as a conscious being (for instance, a tree). within a christian worldview, it can be construed as a sign from god. but for the rational being, consciousness becomes linked with reason (and the assumption here is of one single type of reason, maybe logics?).

i don't think i have a point, just that these things came to mind.

daedalus2u said...

I have a rather long blog which talks about this. I see it as the fundamental trade-off along the autism spectrum, trading off a "theory of mind" for a "theory of reality". A theory of mind is necessary to communicate with others because it allows one to impute the mental state that lead to the communication data stream.


The NT habit of anthropomorphizing everything and attributing consciousness to inanimate objects is a consequence of their over active theory of mind. They impute communication and consciousness when it isn't there, a type 1 error.