7 Ostension and Prehension

In the previous chapter I made a comparison between gripping and referring. It may seem that this is more metaphorical than literal; after all, you do not (usually) physically grip the thing you are referring to. In this chapter I shall respond to this worry by presenting a theory of ostension based on prehension. I shall argue that ostensive pointing can be interpreted as a kind of extended gripping: it is like gripping with an extended virtual hand. It is a kind of imaginary gripping. Gripping is the preadaptation that imagination augments into ostension.

Consider the pointing gesture: the index finger protrudes stiffly toward the object of reference, while the thumb (including its fleshy base) closes over the remaining fingers, which are tucked firmly into the palm. The result is a partially open grip, with one finger extended and the rest closed as if holding something. We do not point with a fist using the index finger’s knuckle as pointer; nor do we open up the fingers and orient them all in the direction of the ostended object. The fingers are partly open and partly closed. What are we to make of this?

We can grip something in the hand so as to conceal it, in which case the fingers close over the object, hiding it. Or we can open up the grip to reveal what is in the hand, partially or totally; a semiclosed grip is advisable if the object might drop or spring free.¹ We can call this the display grip: it is useful in showing to others what is in your hand. The act of showing is an unfurling of the fingers, wholly or partially. We can envisage early humans using a display grip when showing food held in the hand to infants—nuts and berries, perhaps. Closing the fingers and palm over the object would be typically used in the carriage grip—the grip used when transporting objects from place to place. There is also the grip we use when giving an object to someone else, where the identity of the object is revealed, but the object is still partially gripped to avoid slippage. Call this the transfer grip. Imagine an ancient human using the transfer grip to give food to an infant or partner, with the recipient observing the form of the grip. This will be typically a semiopen grip, somewhat like the pointing gesture, in fact. But suppose the parent does not actually have the food in her hand, yet wants the infant to take the food. She might physically touch the food she wants the infant to take, in a kind of lazy gripping action, instead of picking it up. Instead of grasping the fruit and handing it to the infant, she briefly encloses it between thumb and forefinger, indicating that she wants the infant to eat the fruit (she is tired of putting it in the child’s mouth herself and wants him to learn how to feed himself). In so doing she directs the infant’s attention to a particular object in the environment. Or she might just push it toward the infant using her fingers. She uses what we might call a diminished grip. Thus the robust transfer grip naturally evolves into a more gestural diminished grip.

But we can imagine a further diminution or attenuation: the fruit is too far away to touch with her fingers, so the mother picks up a stick and uses it to touch the fruit in question. Maybe in the past she has used a stick to impale the food and physically insert it into the child’s mouth, so this is not much of a departure. Now we have a prosthetic diminished grip. Still the child gets the message: “Eat that piece of fruit!” But what if there is no stick around? The mother might then extend the index finger in the direction of the fruit, so that it is as if she is touching it. Finger and fruit might be only inches apart, or the distance might be greater, depending on the context. After a while the gesture might become stylized, with traces of the original holding grip still present, but the index finger extended out as if to touch the object. The gripping has become virtual—merely imagined. Thus the transfer-display grip has turned naturally and gradually into what we think of as the pointing gesture. It has become a notional gripping, a virtual prehension, an imaginary touching.²

Something of the evolutionary history of the ostensive gesture, so conceived, apparently survives in current practices. On occasion we are not content with distant pointing, perhaps because of possible ambiguity; we might then walk right up to the object and jab our forefinger into it. Thus the army sergeant approaches the delinquent cadet and jabs his forefinger into the lad’s chest, exclaiming, “You, yes you, report to the captain at once!” Or one might walk down a line of people all wishing to be chosen for a team and lightly touch a subset of them with the tip of one’s finger, thereby indicating who is to join the team. We thus reverse gestural evolution, returning ostension to its grip-and-touch origins. It would also be possible to briefly grip each candidate’s shoulder, as a gesture of inclusion in the team—this would serve much the same purpose as pointing the person out from a distance. Also, we often resort to the use of a “pointer” when engaging in certain kinds of ostensive reference—a specially made rod used to touch, say, particular parts of a map about which one is discoursing. Distant pointing with the finger may be too crude and ambiguous, so we extend the pointing finger with a suitable prosthesis. Again, the act of reference has strong connotations of touching and gripping (the pointer itself is gripped).

There is thus a continuous sequence from gripping in the hand to displaying with the hand, to transferring, to touching, to using a pointer, to pointing from a distance. Preexisting forms of hand action are gradually modified to serve a felt need. Ostension has its preadaptations and precursors in antecedent manual acts.³ The shape of the ostensive gesture, as a partially open grip, is testament to its history—we have here an instance of ancestral preservation. As a matter of logical possibility, a species might point with the elbow or the tongue or even the rump; but the way humans point reflects the roots of the gesture in prehensive acts of the hand. Human pointing is prehension modified.

We can also note that the ostensive gesture characteristically involves extending the arm; we do not point “from the hip.” Thus the action simulates reaching for something, which is a prelude to seizing and grasping it. So the whole action looks like attenuated reaching—truncated reaching, we might say. Ostensive reference to an object resembles reaching out to the object. If we had indefinitely extensible arms, capable of reaching and seizing any object in the environment, we could dispense with remote pointing, and simply extend our hands to touch the object of reference.⁴ We can certainly imagine a species with special tentacles that never points without actually touching the object of reference, even gripping it with suction pads for the duration of the act of reference. Our fingers are comparatively short and inflexible, so we replace touching with virtual touching, gripping with virtual gripping. The imaginary line projected from the tip of the pointing finger to the object acts like an extension of the finger, but we can conceive of fingers that never need such virtual supplementation. What we call ostensive reference is touching gone virtual, according to this way of looking at things.⁵

It is instructive to compare the pointing gesture with the so-called come-hither gesture. When beckoning someone, it is common to pull the extended fingers back into a semigrip repeatedly. Compare this with actually seizing the person and gripping him. The come-hither gesture mimics gripping another person and pulling him in. We can see how it might have developed from actually seizing someone, becoming attenuated and stylized. It comes to symbolize what it used to do physically, and is interpreted that way. A person not wishing to accede to the beckoning gesture might extend an upraised palm in a kind of pushing motion, as if resisting the other’s attempt at gripping. We have virtual seizing and virtual resistance. Another example might be the “farewell gesture”—waving goodbye. This action looks a lot like stroking or grooming, and we can easily imagine that it arose by attenuation of actual stroking. It becomes a ritualized gesture of affection or solidarity.

In the light of these examples, we might postulate a law of gestural attenuation: the tendency of a hand action to become modified into a weakened or diluted version of its original, thereby becoming more symbolic than actual. We thus get “action at a distance,” as the original action is performed at some distance from the object of the action, in stylized form. If this is on the right lines, then the ostensive gesture is one instance of the law of gestural attenuation—it results from a weakening of an original action of gripping and touching. In all these cases the derived gesture morphologically resembles an action of the hand that involves actual contact. It is a bit like the “air kiss”: it resembles a real kiss, but is performed at a safe distance. The “fist pump” is similar: at close quarters it would be a punch in the stomach, but from a distance it signifies victory or superiority or domination. The actions become physically detached from their object, but they mimic the original action quite closely.⁶

Thus reaching and gripping become modified into virtual prehension, and this is what the ostensive gesture amounts to. Since the ostensive gesture is a (or the) basic mode of reference, we can begin to see how reference evolves from more primitive prehension (combined with the other factors I mentioned, such as imagination). Pointing to something is a modification of a preexisting form: the action of gripping (and therefore touching) an object. If we think of the original grip as a precision grip, the kind used in holding a nut, say, then the index finger is already assuming the posture of pointing in embryonic form: the nut will be delicately gripped between thumb and partially extended forefinger, with the tip of the forefinger pressing lightly on it. Releasing the grip, as in handing the nut to someone, involves extending the forefinger slightly, so that it becomes straighter. This is the posture of the finger that is used in merely touching the object, as when the mother indicates to the child what to eat; and then increasing the distance between finger and object has the form of the pointing gesture. The transition is smooth and natural: there is no sudden saltation whereby a chance mutation gave rise one day to an animal with an unprecedented ability to point at things. Pointing has its evolutionary prehistory, its precedents and precursors, its embryonic form—following the general laws of ancestral preservation and incremental adaptation. Just to have a convenient (and catchy) label, we might call this the haptic theory of reference, because of its emphasis on the role of touching in the genesis of reference—though this oversimplifies the full story.⁷

Let me end this chapter with a methodological disclaimer: what I have suggested here, as elsewhere, is highly speculative. I have no direct empirical evidence that anything like the story I have told actually occurred. It is also only the beginning of a complete theory. The point is that such a story renders the evolution of ostension intelligible—it enables us to see how ostension could have evolved. This is what may have happened. That is, it is a theory—a hypothesis. I am not asserting that I know this theory to be true, only that it might be true. It is coherent, explains the data, and respects well-motivated constraints on evolutionary explanation. It is a candidate for truth. And I don’t know of any better theory. That is all I claim—not some sort of magical insight into what happened in the remote past, or hard fossil evidence demonstrating that pointing evolved from prehension. (I make this disclaimer in response to stern “scientific” critics who accuse me of claiming more than I am evidentially entitled to, as if I am indulging in some sort of methodological delinquency. That is to miss the point of what I am doing.)