Fifty-six years ago I wrote a brief article for MIND, a philosophy journal edited by Gilbert Ryle at Oxford University. The article is a critical response to Norman Malcolm’s theory that “no amount of intelligible sounds coming from an oak tree or a kitchen table could create any probability that it has sensations or thoughts.”
 |
| Norman Malcolm (1911-1990) |
Here is a summary of what I wrote.
Malcolm’s supporting example is a talking tree. The tree “says” that “there is a vixen in the neighborhood.” And then, on request, the tree not only “defines” the word “vixen,” but it also defines the words in the definition, i.e. the words ”female” and “fox.”
[I put quote marks around “says” and “defines” because these words imply that the tree is speaking.. What I think Malcolm should have said is, “The words ‘There is a vixen in the neighborhood’ appear to come from that tree.”]
Getting back to Malcolm’s claim that there is no probability that the tree or table has sensations or thoughts, Malcolm gives the-example of a child who correctly uses the word ‘red’ by ‘picking out’ things that are red, e.g.a red ball. The child looks at, points to, reaches for, and goes to, the red ball, not the green or blue ball.
Malcolm writes that the tree is logically incapable of such behavior. It does not have the human form. Without the human form, a thing like a tree, table, or computer, cannot point to objects like a red ball. It cannot understand the words that emanate from it. It cannot make ‘correct applications’ of the words “There is a vixen in the neighborhood.”
In my paper, I have two responses to Malcolm. marked (a) and (b).
First, to be clear, according to Malcolm, a necessary condition for picking out objects to which one’s words refer is having the ability to point or look at, fetch or get the object. If you don’t have these abilities, then you can’t perform the activity of ‘picking out.’
In part (a), my counter example is a color-blindness test. If you are asked to identify the red ball from a distance you can do this by saying, “It is the ball on the left.” My point is that we can pick out things verbally (vocally, audibly). I call this ‘acts of verbal indication’. If we add ‘verbally indicate’ to the list of ways that one can pick out an object to which one’s words refer, then this is a communicative activity that (logically) can be done by beings that do not have the human form.
Going back to the vixen example, suppose the words uttered by (emanating from) the tree are, “There is a vixen standing behind you.” The tree has 'picked out' the location of the vixen by voice alone. Since Malcolm has already used the words ‘says’ and ‘defines’ when describing the activity of the tree, then why not also use the words ‘picked out’ as another activity? Perhaps having the human form is not a necessary condition for picking out objects referred to verbally.
What I am trying to do in part b is to show first, that human beings are not the only things that have the human form. Consider, for example, artificial things like Barbie dolls, puppets and robots. How would Malcolm distinguish between human action and the movements of artificial things that have the human form?
I ask this question because we often anthropomorphize our language when describing the movements of inanimate things, whether or not the inanimate thing has the human form. The puppet “walked” across the stage, but also: the hurricane “threw” the car into the lake. The tree “dropped” its fruit to the ground.
So what is in common to these descriptions? One commonality is that we cannot use the adverbs ‘intentionally’ and ‘unintentionally’ when talking about their movements. ‘The tree dropped its fruit’ is identical to ‘the fruit dropped from the tree’. We would not say ‘The tree intentionally dropped its fruit,’ nor ‘The tree unintentionally dropped its fruit.’
The same thing would be said of the movements of the puppet or the robot. Their movements are neither intentional or unintentional, voluntary or involuntary, deliberate or not deliberate. But the puppet master can unintentionally do something that affects the movement of the puppet, and the programmer can unintentionally do something to the robot’s program that affects the movements of the robot. [Compare programming an electric automobile]
My conclusion is that it is intentionality not the human form that distinguishes human action from the movements of such things as trees, kitchen tables, puppets and robots. The tree can’t point or look at, fetch or get, not because it lacks the human form but because the tree (and the robot) cannot do these things intentionally or unintentionally, deliberately or not deliberately, voluntarily or involuntarily.
And if it is protested that intentional behavior is only possible in those animals that have the human form, then tell this to a dog owner or to the aliens from outer space who look like lobsters when they walk down the ramp from their space ship while shouting “Humans look good to eat.”
Can intentionality be applied to AI bots? Can a robot fueled by AI make a mistake or do something unintentionally?
I’ll leave that question to you.
_________________________________
Houlgate, Laurence D. “Malcolm on Mind and the Human Form,” October, 1968. MIND. VoL XXVII: 584-587.
Malcolm, Norman. Knowledge and Certainty: Essays and Lectures. 1963. Prentice-Hall.
