On the argument from silence

I have finally finished reading C.L. Hamblin’s Fallacies, after putting it down and taking it up on at least eight different occasions over three weeks. I was already familiar with it as a seminal text in argumentation and informal logic, having had the late Michael Leff summarize its findings in depth in a class two years ago, but I have special occasion for digging into it in detail now. I’ve been turning over an idea concerning fallacies for a few months, and having followed with interest a recent running debate of sorts between Vridar and Exploring Our Matrix on the question of the existence of a historical Jesus, I think I’m onto something.

So I’ve started working on a piece, set within a certain debate in biblical studies over Paul’s epistles, that analyzes the bad reputation of the oft-charged ‘argument from silence’ fallacy (also known informally as the AOS, or more formally as the argumentum ad ignorantiam or argumentum ex silentio) and argues that it isn’t a logical fallacy at all; rather, it is only a dialectical rule or guideline (as are most of the “logical” fallacies). Thus, the vast majority of the time someone cries ‘AOS fallacy!’ they are invoking not logic, but a violation of arbitrary argumentative protocol that itself depends on who holds the burden of proof (itself a matter of popularity). This would be of no surprise to those familiar with informal logic (a field that seems to have more or less been brought into being by Hamblin, Perelman, and Toulmin), but it might open an eye or two in biblical studies, which contains competing epistemologies that often invoke the AOS for different purposes.

A hypothetical example showing the various ways the AOS is often used (or decried) may help unpack a little of what I’m thinking about lately. I’ll avoid the actual debate on Paul that I want to explore, to show these kind of argumentative maneuvers are universal.

Let’s say a rhetor (Bob) and another rhetor (Sue) engage in a discussion (or dialectical exchange) about the existence of UFOs.

Bob begins by claiming that UFOs probably (remember this key adverb) don’t exist, because there is no positive scientific evidence for their existence: no wrecked ships, no unambiguous photographs or telemetry, alien bodies, productive chats with little green men on the White House lawn, etc, etc.

Sue counter-claims by charging Bob with AOS (argument from silence) usage. While she freely admits no positive evidence exists, she points out that no negative evidence exists either; the lack of UFO evidence does not logically rule out their possible existence. Furthermore, she adds, thousands and thousands of people have claimed UFO sightings; even if the vast majority of them were mistakes or hoaxes, it only takes one positive case (remember this, too) to establish UFO existence.

Bob counter-claims by charging that Sue, not him, is using an AOS. If he accepts her terms, it would be impossible for him to ever prove that UFOs don’t exist; he would have to ‘prove a negative,’ which is impossible. It is upon Sue, he states, or any other person who wants to show the probable (again, remember this word!) existence of UFOs, to bring scientific evidence forward. The sheer number of false positives is strongly indicative of all future claims to also be false positives. Therefore, it is reasonable to hold that UFOs probably don’t exist, and even to say that it is highly probable or even effectively certain that they do not.

Sue, nonplussed, states that Bob’s point about probability is well taken… but notes that if Bob is going to uses probability as his standard, then he should keep in mind that any number of false positives among sightings guarantees nothing about the state of future sightings; this is the same fallacy that led to the Challenger disaster (all the previous shuttle launches had succeeded, but the chance of failure did not decrease from launch to launch). She also notes that as humans have launched spacecraft before – and even if we were still on the cusp of doing so – there is no theoretical barrier to another intelligent form of life doing the same. Sue further notes that discounting the possibility from debate artificially lessens the chance that any evidence favorable toward UFO existence will ever be interpreted in that light, or freely discussed; Bob’s stance eliminates a real possibility before it can be safely dismissed.

Bob is getting a little frustrated by this point with Sue’s tenacity. He notes that Sue’s reasoning could be applied to virtually any idea or concept, including the Easter Bunny, Santa Claus, and the Greek pantheon. Maybe it is “possible” for the Greek gods to be real (as they are in the Percy Jackson movie about to come out) but the possibility of positive evidence for such appearing is very low – so low, in fact, that seriously considering the idea is a waste of time. Sue, he charges, has a poor grasp of the possible, so much so that she has likely confused it with probability.

Sue’s counter is quick. She notes that the question of UFO existence is very different from his three extreme examples. For one, UFOs are (or should be) physical objects with no special mythical or magical properties, obeying known physical laws; they need no special exceptions to exist. Bob’s position, on the other hand, requires that no intelligent life in the galaxy exists – or if it did exist, that it has not developed space travel or at least interstellar space travel – and if that is granted, then Earth has never been visited. With not a small amount of glee, Sue notes that this is a huge, chained example of an AOS – so extended, in fact, that it is Bob, not her, that seems confused as to what is possible.

I could go on, but I will stop here, and note that neither Bob nor Sue has used much formal logic to make their respective cases. Syllogisms are next to non-existent in even scholarly arguments, anyway. Rather, this is the kind of arguing that results, quite normally, from any question that has no clear, definitive answer, given a starting point of insufficient present evidence, evidence in open conflict, and/or strongly held assumptions. In other words, the question has stumbled into the realm of rhetoric, which I’ll define as the art of convincing others to make decisions when certainty is not available due to a lack of evidence, conflicting evidence, or strongly held, resistant assumptions (I’ll note that I have not gotten into the question of whether or not Bob or Sue is biased, in that they may have a personal stake in the UFO question – this is extremely relevant in terms of being convincing, but has very little to do with formal logic or having the better theory).

Entering the realm of rhetoric tends to make scholars concerned with certainty very uncomfortable, perhaps as they can sense that certainty is being manufactured rather than established. This leads to the creation of rigorous criteria that supposedly establish objectivity and banish subjectivity. But that’s the irony, because all certainty, certain or not quite certain, is manufactured. The rhetoricians looking into the rhetoric of science have pretty much nailed the epistemic nature of scientific discourse to the wall – though, of course, I won’t say they’ve done so certainly, or that I have.

Analyzing the above pseudo-debate, Bob is very much concerned with what is probable, and feels comfortable excluding certain hypotheses or possibilities from the table; he knows certainty is impossible, but he would like to be right, and his current hypothesis, in his view, increases his chances of being right dramatically. Sue, on the other hand, is more concerned with possibility, and has a dimmer view of the reliability of human judgments of probability; she, too, would like to be right, but excluding possibilities to her seems reckless, even closed-minded.

What we have here is not a logical impasse full of corresponding fallacies, then; what we have here is a fundamental disagreement about the best way to advance knowledge and understand the world. One forms high-probability hypotheses and concentrates on developing them at the expense of low-probability hypotheses; the second, distrusting the consensus of what is probable as well as the possibility of accurate measurement, concentrates on presenting a full range of possibilities. If you hold to Sue’s epistemology, it’s very acceptable, even ethical, to not discount low-probability possibilities if they are theoretically possible. If you hold to Bob’s epistemology, though, it’s a waste of his time.

As Bob’s way of thinking is generally predominant in the scholarship of many fields, due to its efficiency, the AOS has a bad reputation, even though his epistemology, as Sue charges, requires its usage from time to time. When there is an absence or rarity of evidence for some questions, the AOS creeps in on the most “rigorous,” as there are a surprising number of cases where it can be used very reasonably (which I’ll leave to a later post).

Now you may have guessed that I am somewhat sympathetic to Sue’s position here, but not entirely. I think, rather, that these two epistemologies are not necessarily in conflict.

Bob’s epistemology is great for summarizing the current views of the field on a given question, for teaching, and in general, for deciding between competing theories. Its conservative nature, however, makes it slower to embrace difficult or unpopular answers to questions, and harder to spot fresh possibilities in well-explored topics, as all theories must survive merciless gauntlets of criterion designed to exclude low-probability theories.

Sue’s approach, however, is better for making paradigm-changing discoveries (insert obligatory Kuhn reference; perhaps I should have done this earlier when talking about criterion, actually). She will be far more likely to refuse to pronounce a topic or question dead; she will not shy away the AOS, but use it as a tool like other criterion to uncover unconsidered or prematurely dismissed possibilities; she will be less intimidated by consensus; she will more cheerfully attack established criteria. The danger, of course, is that she will find it far, far harder to get her ideas considered, even though she has broken no logical rules and practices rigorous consistency. It is much, much harder, in more than one way, to be a quality Sue than a quality Bob.

The best scholars I’ve encountered seem to mix a good working sense of the probable with a tacit acknowledgment of the possible, understanding that entertaining all possibilities on a question is not the most profitable use of their time, but also keeping a lookout for the overlooked, misinterpreted, or ignored possibility. Central to this balance, I think, is some flexibility in regards to questions that lead into the AOS. More later.

Leave a Reply