Certainty as Chance

It is clear, for a philosopher, that some kind of certainty of claims is vital. Yet in one paragraph, David Hume seems to declare all of philosophy to simply be quibbling over instances and hoping that truth will arrive by accident: “...`tis impossible for us to satisfy ourselves by our reason, why we should extend that experience beyond those particular instances, which have fallen under our observation.” What is the evidence for this claim?

Assume, for a moment, that one could discover a completely self-evident cause-and-effect relationship in the past, such that one is absolutely confident that an apple released from a man's grip would fall to the ground or that a particular trajectory of billiard ball would propel a different ball to a different trajectory (or, to use an example discussed in lecture on October 20th, that a certain fermented concoction in a chilled glass would instill a feeling of warmth and freedom). Even this will not do, Hume says. For, as he describes in T 1.3.6.12, “ Your appeal to past experience decides nothing in the present case; and at the utmost can only prove, that that very object, which produc'd any other, was at that very instant endow'd with such a power; but can never prove, that the same power must continue in the same object or collection of sensible qualities; much less, that a like power is always conjoin'd with like sensible qualities.” Even if we are totally confident of one instance of cause-and-effect that we can identify as not mere coincidence, or even an infinite array of past instances, we have no reason to expect based on empirical evidence or the story of our senses and memories that that power will be retained into the future. The rules of the universe, or of a localized space, may change completely tomorrow. Attempting to establish with evidence from the past that evidence from the past is reliable, a classic scientific response, is viciously circular; as Hume indicates, “If you answer this question in the same manner as the preceding, your answer gives still occasion to a new question of the same kind, even in infinitum; which clearly proves, that the foregoing reasoning had no just foundation.”

In addition, as Hume declares in 1.3.6.15, “We cannot penetrate into the reason of the conjunction”; the causal relationship is uncertain because it is a human conjunction. While one may see event X followed by event Y, it is only the provision of the mind that generates the causal relationship. To be fair, Hume almost seems to establish this through definition, moving through his fork until it becomes clear that of course a cause cannot reside in an impression and must be an idea that is formed. Yet this does beg a question: If one lights a fuse of a bomb, is there not a cause that is clearly visible that one can see? Or is the cause rather the chemical interaction of the detonation? Nonetheless, the point is reasonable to accept. Though Hume does not have the language to make this claim, one can bear in mind the admonition in statistics to avoid confusing causation with correlation, indicating that causation requires something more than simple consistent connection between two vectors.

To analyze properly, one must separate the cognitive implications of Hume's argument from the epistemological import. Hume proposes not only that constant conjunction is an empirically uncertain way to proceed, but that (strangely enough) it is what humans automatically assume. In 1.3.6.15, he proposes, “We only observe the thing itself, and always find that from the constant conjunction the objects acquire an union in the imagination.” One wonders: If Hume is such a skeptic, how is he willing to make such concrete claims about human cognition? Putting that aside for a moment, there are a few ways one could test the hypothesis. It seems some danger responses are ingrown: no matter how uneducated the man, he will always duck a fastball; a baby will not cross a glass pane unless he has good reason to believe it will be safe1. It seems there are certain reflexes that are inborn. Yet it is also obvious that a man who has not been told that a particular liquid will be intoxicating or that a particular small white ellipsoid shape will help cure some malady will proceed as comes naturally to him without that information: perhaps drink the liquid without checking, perhaps smell the liquid and taste it thereby deciding not to drink it, perhaps not take the pill because most people don't put random objects into their mouths. With that information, they will proceed differently. More, they are more likely to trust that the above results will be inevitable if they are told by a trusted friend or professional. Here, the causal chain is, “X actor that has given good results before tells me to do Y thing. Since good results have always emerged from doing what she tells me to do, Y will also have good results.” (To reaffirm the causal questions above: Epidemiology tells us that some drugs are allergic to a minority of the population, such that the person above might be mistaken in making that assumption; and that many medicines work to varying degrees, sometimes not at all; and often inaccurate diagnoses of either symptoms or causes occur such that the drug will be ineffective). If the person drinking the liquid or taking the pill then discovers that the results are pleasant, they may continue to do so in a variety of different circumstances, even without guidance or with trusted friends; and, alternately, if the results are poor, no amount of trust may elicit a repetition of the result.

Does Hume mean that this response occurs involuntarily, such that every time one sees or somehow confronts X external influence it becomes associated with Y idea or impression? Evaluating this claim is difficult, of course, but it should be obvious that people forget some events such that they thereby will not draw a cause-effect relation, or will only vaguely remember, or will remember competing things, or will dismiss the associated memories as irrational or coincidence. Indeed, what is often called an “open mind” is precisely the ability to suppress this bias of linking two events. One may have bad memories from a particular ethnic group, such that encountering a member of that group makes one imagine a causal relationship wherein an unpleasant result will occur, yet one will (perhaps on the advice of friends or family) try to ignore that association. Yet this may be unfair to Hume. As indicated above, Hume is not necessarily speaking of the rational inference through induction of causal relationship; he would view those as intellectual constructions of convenience. Rather, he is saying that the human mind does not imagine something without connection to other referents and, more importantly, that “'tis an idea related to or associated with a present impression”. One would clearly not assume a theory of gravity if one regularly saw apples floating in mid-air. The more times one sees a particular set of circumstances being conjoined and the more clear the conjunction, the more likely the inference of a causal relationship, even if that relationship cannot be spelled out. If a man sees a black cat and then is struck by lightning once, he may view it as a freak coincidence. If that number were ten, or one hundred, he might view the situation somewhat differently.

To refine the above: Consider the matter of the billiard ball. Hume alleges, in the eleventh paragraph of the Abstract, that “Were a man, such as Adam, created in the full vigour of understanding, without experience, he would never be able to infer motion in the second ball from the motion and impulse of the first.” This may or may not be true. A person of a more Chomskyan persuasion would propose that there might be and probably is a natural instinct of humans to track motion and anticipate its result. But even if one assumes this result, it does not go very far in disproving Hume. Hume, in fact, was setting out to produce a science of human nature. The fact that cognitive biology and psychology may complete his vision would appeal to him, not irritate him. One must also remember Hume's assumption of a Lockean framework, including the tabula rasa principle. Now, assume Adam decides to become a novice billiard player, with some degree of talent. He may have a vague natural notion of the motion of objects. But he would have no reason, when beginning the game, to assume that the balls weren't too heavy to move with a wooden stick, or perhaps were unfairly weighted to lean to one side or the other. After enough repetitions, as Hume says in the 12th paragraph of the Abstract, “If he [Adam] had seen a sufficient number of instances of this kind, whenever he saw the one ball moving towards the other, he would always conclude without hesitation that the second would acquire motion.” Over time, he would refine this notion, acquiring a more and more precise understanding, perhaps even intuition, of the way billiard balls operate, of how to coordinate his body and his stick to produce desired velocities and trajectories, of how to avoid “scratches”, and of all the other necessary faculties to play a good billiards game. He would do so through a process of more and more sophisticated cause-and-effect relationships that would indeed become so ingrained as to be almost automatic.

What does all this mean for humanity? Is it doomed to proceed blindly with no certain knowledge whatsoever? Or is it to use reasonable certainty, in line with its unique cognitive abilities, to survive in a complex world and satisfy its ethical and personal passions? Hume will take this up later on in his work.

Sources:

Gibson and Walk, 1960. The experiment was called “the virtual cliff”, wherein babies were placed on one table, a glass pane set across to another table, and a mother at the other table called the baby. Infants regularly refused to cross the pane. This is why many psychologists argue that the only two innate fears are of heights and loud noises. http://www.learningmatters.co.uk/sampleChapters/pdfs/cognitive%20chap%202.pdf for more.