Arguments from Silence
This is one of two principles that are often cited in Latin. The Latin phrase for this one is
argumentum ex silentio
or the "argument from silence." We may here notice how this phrase is used in applied history, and then consider what else may be implied when something is not mentioned in the texts.
"Silence" means that the thing in question (call it X) is not mentioned in the available documents. If it were mentioned, then with the usual qualifications it would be proved to exist. Since X is not mentioned, X cannot be proved to exist. A natural further inference from this evidence is that X did not exist. The basic point is that if X did not in fact exist, then the only trace which that fact could leave, in the evidence, is the silence of the evidence as to X. At the same time, any such conclusion must be provisional. If documents are later found that do mention X, then X is after all proved to exist. A single positive may overturn any number of negatives. A single sound refutes all silences.
The possibility of such a future positive can never be ruled out. But until it occurs, the non-existence of X is the best inference from the absence of X in the evidence. The strength of that inference in a given case will depend on (1) how many documents there are, or in statistical terms how large the sample is, and, in literary terms, (2) how likely the thing is to have been mentioned in documents of that type in the first place. We might explore these concepts just a little.
The converse of the first point is that if newly discovered documents continue not to mention X, then the case for the non-existence of X is proportionately strengthened. This is what statisticians call a sampling question. We may take the question of early Chinese swords as an example:
- From statements in Warring States period texts, it was presumed that many Spring and Autumn period swords must have existed. Archaeology failed to provide evidence for them. In the early 20c, when few metal artifacts had been recovered, the absence of significant numbers of swords could be dismissed as a chance result, and the stories in the texts could still be accepted as factual.
- By the late 20c, many more metal artifacts had been recovered. The pattern of the new finds largely confirmed the pattern previously known. There were many new examples of known common types, but no new types were found. At this date, it was conceded that there was indeed a Problem of the Swords. Doubt began to be cast on the reliability of the texts. The problem was formally raised by David Keightley, in a 1976 article called "Where Have All The Swords Gone?" (EC #2, 31f). Sequels followed by William Trousdale ("Where All the Swords Have Gone," EC #3, 65f; doubting that the argument from silence was valid) and Noel Barnard ("Did the Swords Exist?" EC #4, 60f; pointing out the risks of clinging to text-based presumptions which archaeology consistently fails to support).
Even that considered conclusion is technically a working hypothesis. But at some point, a hypothesis from silence properly comes to be seen as capable of bearing weight; of doing work in history.
Arguments from silence can be strengthened by the presence of something (call it Y) which replaces the conjectural X. Thus:
- The Confucius of Analects 4 never mentions or cites the classical texts. So also with the Confucius of Analects 5. So also with the Confucius of Analects 6, and so on up to Analects 9. These are the earliest layers of the Analects, which is agreed to be the best source for the historical Confucius. The implication is that the historical Confucius didn't know, or teach, the classical texts. But these chapters of the Analects are small; perhaps the sample is not decisive? That is certainly a possible option. But on further inspection, we find that Confucius in these chapters teaches in another way: he has his followers meditate on certain maxims of behavior so as to work out their implications, or school their moral sensibilities by observing moral or immoral behavior in others. The absence of the classical texts is as a teaching tool is suggestive, but the presence of a quite different teaching method, rooted in the observable present and not in the past, is somewhere near decisive.
- It was noted, in discussions in the year 2000 on the WSW E-list, that there were no instances of forgiveness in the writings of the Warring States period. That fact by itself has many possible explanations, along with the obvious explanation that forgiveness was not a prominent element in Warring States culture. What tends to support the absence of testimony for forgiveness is the massive presence of testimony for unforgiveness, in the form of the vendetta, where injuries were avenged either directly or by use of hired assassins.
In sum, the argument from silence, like all historical arguments, is always conjectural. But it is not, as some claim, a fallacy. It is the correct default inference from silence. That inference can be strengthened by relevant evidence of a positive kind, or by the continued silence of further evidence.
Étienne Trocmé said it about right, in the Preface to his book The Childhood of Christianity:
There is no reason why new documents should not appear one day. So my conclusions are provisional and are always open to modification. However, it should be noted that the ancient libraries and archaeological sites which since the middle of the last century have given us so many unpublished documents, and very old manuscripts known hitherto only from later copies have now been the object of very thorough investigations. So the probability of new sensational discoveries is low. The account which follows may thus be regarded as relatively trustworthy, to the degree that it interprets the available documents correctly.
It is indeed a probability estimate. A statistical estimate. In certain well-defined situations, there is a precise point at which silence, or nonoccurrence, becomes statistically significant:
- If your friend tosses a quarter and gets 1 Head out of 1 attempt, chance is not excluded. The odds of this result occurring by chance are 1 in 2 (0.500). If he tosses a quarter 3 times and gets 3 Heads, that is, if the Tails result continues not to occur, the odds are reduced to 1 in 8 (0.125) that the coin tosses are honest. But mere luck is still not out of the question. At what point, then, may one validly begin to suspect that more than chance is at work; for instance, that your friend is equipped with a two-headed quarter from the local joke shop? These questions have answers in statistics, and some "silence" situations are amenable to exact calculation of this sort.
You first have to define how certain you want to be, but taking the conventional level of "99% certain," the answer is: after 7 consecutive tosses of Heads, it is 99% certain that something is wrong with the situation. Our advice would be to quit playing at that point. The hypothesis of honesty is no longer tenable.
As always, we need to be aware, not only of the numbers, but of the culture behind the numbers. There are various reasons, other than literal nonexistence, why some item of culture is not, or seems not to be, mentioned in the texts of the time. Such situations do not imply nonexistence.
- The item is too familiar to need explicit reference by members of the culture. Underneath the named virtues in early Confucianism is a second set, not named, which must be inferred from the discourse. One such tacit item is the value of effort. Early China (like modern China) is an "effort" culture. It never names itself an effort culture. Why not? We would suggest: because it was too fundamental to require discussion or even notice. What is discussed in the elite texts are the points at which elite culture differed from the rest of the culture, or (especially) at which different factions within the elite culture differed from each other.
- The item is unlikely to be mentioned in the type of texts which have survived. We have no details of the daily life of ordinary people in the WS philosophical texts, because all the extant texts derive from the elite stratum of society. The preservation mechanism for even those texts is the continued institutional existence of the faction or school out of which they came. No documents of "ordinary" origin exist, even from periods when the non-elite population could read and write. Such documents would have had only private structures for their preservation over time (the author, his relatives), and private structures tend not to persist very long over time. Stuff gets tossed; wars happen. You may possess something that belonged to your father. You are not very likely to possess something that belonged to your great-grandfather, unless he was a person of unusual note, whose deeds and paraphernalia were more strongly preserved, and in that case, they are more likely to have been preserved or commemorated in a museum or a monument.
- The item is culturally taboo. Non-Chinese peoples are mentioned occasionally in texts of the 04th century, but after a certain point, such mentions cease. The point where such mentions stop is probably the point at which hostilities escalate between the Chinese and a new coalition of steppe peoples to the north. The existence of a society comparably organized but adversely disposed was a fact which the Chinese world view could not readily accommodate (we can see it wrestling with the problem in an atypically explicit passage in Analects 3:5). A probable parallel is the near or complete absence of references to writing in early Sanskrit texts. Writing in early India seems to have been a tool of the culturally disprized traders, and to have been scorned accordingly by the Brahmins who were responsible for most of the extant texts.
- The item actually is referred to, but allusively. Early China was a virtuoso culture, where direct reference was considered unstylish, and an authoritarian culture, where direct statements might be dangerous. A a highly allusive and indirect way of mentioning certain things comes to be increasingly prized. There are many people alive who can read with understanding the underground or samizdat literature of Soviet Russia. But the allusions and indirections of a remote and imperfectly known culture may be impossible to decipher, or even to detect.
- The item actually is referred to, but sarcastically. Sarcasm and its cousins don't survive well even within a culture. Some of the most barbed sarcasms of the Analects are taken "straight" by most later readers of the Analects, and they are no less likely to be missed by outside readers.
These are among the cautions and considerations. A serious argument from silence will take them into account. But the argument from silence to nonexistence remains in principle a valid inference; the least unlikely conclusion to be drawn from the facts presently at hand.
17 Dec 2006 / Contact The Project / Exit to Outline Index Page