The argument from silence.

A brief look into a perilous kind of argument.


Refer also to my page on historical methodology.

About the argument from silence.

Historians generally regard the argument from silence as the weakest weapon in their arsenal. The argument from silence attempts to demonstrate that an alleged historical event never actually happened, and it attempts to do so based on the silence of an author or group of authors on that event.

Any single argument from silence, then, stands on two premises both of which must be true in order for the argument to merit the attention of the historian. First, the silence of the author in question has to mean that the author did not have information about the alleged event. Second, this lack of information about the alleged event has to mean that the event did not happen.

Each of these premises can very often be regarded as specious by itself; when, therefore, an argument depends on both of them together, the overall argument can be perilous. The argument from silence, then, must be used with extreme care. It may be helpful to outline some conditions, or tests, for the use of this kind of argument.

The burden of proof.

The burden of proof always rests on the historian making the assertion of fact or nonfact. In the case of the argument from silence, the assertion is one of nonfact, and the one making that assertion bears, as it were, a double burden.

Gilbert J. Garraghan writes on page 162 of A Guide to Historical Method:

The argument from silence aims to prove the non-reality of an alleged fact from the circumstance that contemporary or later sources of information fail to say anything about it. It is sometimes misleadingly called the negative argument; but this can easily be taken to mean something false, namely, that the argument rests on an explicit denial of some fact.

Thus Garraghan aptly reminds us that it is one thing for an author to actually deny a certain datum, quite another for that author merely to pass over it in silence, neither denying it nor affirming it. The former is not an argument from silence; only the latter is.

Garraghan goes on to offer two conditions that an argument from silence must fulfill in order to be used in an historical argument:

  1. The author withholding the alleged information was in a position to have that information.
  2. The author withholding the alleged information would have certainly made mention of it had he or she known of it.

Note that these two conditions bolster the two premises put forth above for any argument from silence, to wit, (A) that the author should have possessed the information and (B) that the author should have divulged the information. This is the double burden of the argument from silence.

After using an argument from silence in an example involving Israeli involvement in the 1982 attack on refugee camps in Beirut, Martha Howell and Walter Prevenier write on pages 74-75 of From Reliable Sources:

Of course, an argument from silence can serve as presumptive evidence of the "silenced" event only if, as in this case, the person suppressing the information was in a position to have the information, and was purposing to give a full account of the story from which he omitted the crucial information, and if there were no compelling reasons why he should have omitted the information (other than the wish to conceal).

Howell and Prevenier, then, propose three conditions for the proper use of an argument from silence:

  1. The author withholding the alleged information was in a position to have that information.
  2. The author withholding the alleged information intended to give a full account of the event from which he or she omitted that information.
  3. The author withholding the alleged information had no compelling reason to have omitted the information.

Garraghan agrees with Howell and Prevenier on the first condition, that the author in question was in a position to possess the information, if it existed. Garraghan has no condition to match the second put forward by Howell and Prevenier, that the author intended to give a full account. Finally, the second condition for Garraghan lines up with the third for Howell and Prevenier, but with different wording; where the former says that the author would certainly have offered the information had he or she known it, the latter say that the author had no compelling reason to omit the information.

It is, in my judgment, Garraghan who has counted his conditions most appropriately. The second condition that Howell and Prevenier adduce should not really be a condition at all. For, while it can well be said that if an author intended to give a full account any omitted information is an indication that the author did not possess that information, it cannot equally well be said that only an author intending to give a full account can offer a telling silence. As Howell and Prevenier appear to concede with their third condition, sometimes an author can be practically compelled to relate a particular datum if he or she knows it, regardless how full an account that author intends to give in other respects.

Furthermore, Garraghan is more correct in his second condition than Howell and Prevenier are in their third. To have no compelling reason to omit a particular fact is not necessarily to be compelled to state that fact. Sometimes an author is negligent or forgetful. Rather, as Garraghan puts it, it is up to the person formulating and using the argument from silence to show that the author would certainly have mentioned the datum in question, and that therefore his or her silence on the matter is an indication that he or she did not know about that datum.

The burden on the historian who wishes to invoke the argument from silence, then, is heavy. Such an historian must show both (A) that a given author should have known about the disputed event (if it really happened) and (B) that this author should have written about it (if it really happened and he or she knew about it). The burden on the historian who wishes to ignore the argument from silence, however, is relatively light. Such an historian has only to show either (A) that the author in question may not have known about the disputed event or (B) that the author in question was not compelled to write about it.

Let me briefly note what one might call half an argument from silence. It is possible to argue to the second condition alone, to contend only that the author, based on his or her silence, did not know of a given event, not necessarily that the event never happened. This is a more modest argument, of course, than what I am calling the argument from silence in this piece; nevertheless, it still frequently suffers from the intangibility of what an author, based on his or her knowledge, should or should not have mentioned. The fact is that, of the two conditions, it is the second (the assumption that the author would have mentioned a datum had he or she known it) that is the more perilous of the two. In my Pauline examples below, it will be the second condition that leads us astray.

Three inconsequential Pauline silences.

We might best illustrate the perils of the argument from silence from a trio of presumably noncontroversial examples from the epistles of the apostle Paul. The fact that Paul is silent on so many details of the career of Jesus of Nazareth has led many to propose that many of those details are in fact inventions that postdate Paul.

It is not my present purpose to examine any of the usual silences regarding the career of Jesus; as I have mentioned, my examples will be noncontroversial so as to clearly demonstrate the perils of the argument from silence. In the first section I will make the case, based on the argument from silence, that Paul did not know three particular dominical sayings; therefore those three sayings are spurious. But then in the second section I will show why each argument from silence fails.

The three silences.

1. In Romans 7.1-4 the apostle Paul bases an argument for the mystical death and resurrection of the saints on the precepts of the law:

Or do you not know, brethren, for I am speaking to those who know the law, that the law has jurisdiction over a person as long as he lives? For the married woman is bound by law to her husband while he is living, but if her husband dies she is released from the law concerning the husband. So, then, if while her husband is living she is joined to another man, she will be called an adulteress; but if her husband dies she is free from the law, so that she is not an adulteress though she is joined to another man. Therefore, my brethren, you also were made to die to the law through the body of Christ so that you might be joined to another, to him who was raised from the dead, in order that we might bear fruit for God.

If the married woman is bound to her husband as long as he is alive, then divorcing or leaving her husband to be joined with another man must constitute adultery. But what is missing in this discussion, then, is any mention that Jesus had spoken words against divorce and adultery during his ministry. In Mark 10.11-12, for example, Jesus teaches:

Whoever divorces his wife and marries another commits adultery against her, and if she herself divorces her husband and marries another she commits adultery.

There are parallels in Matthew 5.32 = Luke 16.18; Matthew 19.9. Why does Paul not employ this powerful dominical affirmation of the lifelong bond of marriage alongside his description of what the law teaches? Perhaps Paul did not know this dominical saying. Since he was in a position to receive these kinds of dominical commands, perhaps this particular dominical command did not yet exist in his day.

2. In Galatians 6.6 the apostle Paul writes:

The one who is taught the word is to share all good things with the one who teaches.

This principle is a natural corollary to that which Jesus enunciated to his apostles when he sent them on their mission to Israel during his ministry. Take Luke 10.7, for example:

But remain in that very house, eating and drinking the things from them. For the worker is worthy of his wage. Do not keep moving from house to house.

There is a direct parallel in Matthew 10.10 and an indirect parallel in Luke 9.4. Paul clearly intends to motivate the Galatians to share material things to those who teach the word to them. Why then does he not employ the dominical principle that the worker is worthy of his wage? Perhaps Paul did not know this dominical saying. Since he was in a position to receive these kinds of dominical commands, especially one such as this which would have been so relevant to his apostleship, perhaps this particular dominical command did not yet exist in his day.

3. In 1 Corinthians 15 the apostle Paul argues against certain Corinthians who have apparently denied that there will be a resurrection of the saints. He writes in 15.12-13, for instance:

Now if Christ is preached, that he has been raised from the dead, how do some among you say that there is no resurrection of the dead? But, if there is no resurrection of the dead, not even Christ has been raised.

In the effort to reteach these Corinthians that the saints will, like Jesus, indeed be resurrected from the dead, Paul ignores the various dominical sayings in which Jesus himself affirmed that there would be a future resurrection. Jesus says in John 6.40:

For this is the will of my father, that everyone who beholds the son and puts faith in him will have eternal life, and I myself will raise him up on the last day.

Refer also to John 6.39, 44, 54; 11.24. In Matthew 22.23-33 = Mark 12.18-27 = Luke 20.27-40 Jesus also teaches about gender and marriage in the resurrection. But nothing in 1 Corinthians 15 even hints that Paul knew that Jesus had taught about the resurrection. Perhaps, then, Paul did not know this dominical saying. Since he was in a position to receive these kinds of dominical commands, especially one as important as the doctrine of the resurrection, perhaps these particular dominical sayings did not yet exist in his day.

The three refutations.

1. Paul fails in Romans 7.1-4 to refer to the dominical teaching on adultery and divorce, but he does not so fail in 1 Corinthians 7.10-11:

But to those who have married I give instructions, not I but the Lord, that the wife not depart from the husband; but, if she does depart, let her remain unmarried or let her reconcile with the husband; and let not the man put the wife away.

Paul knew the dominical command on divorce and adultery* after all. The argument from silence, as applied to Romans 7.1-4, would have led us astray on this point.

* 1 Corinthians 7.10-11 does not mention adultery per se, but rather focuses on not divorcing. But the point remains that a dominical command not to get divorced would be relevant to the principle that a woman is bound to her husband as long as he lives.

2. Paul fails in Galatians 6.6 to refer to the dominical teaching on the worker and his wage, but he does not so fail in 1 Corinthians 9.14:

So also the Lord directed those who proclaim the gospel to get their living from the gospel.

Paul knew the dominical command on supporting the teacher* after all. The argument from silence, as applied to Galatians 6.6, would have led us astray on this point.

* 1 Corinthians 9.14 approaches the matter from the point of view of the apostle or teacher, while Galatians 6.6 approaches the matter from the point of view of the one taught. But the point remains that, if the Lord intended apostles or teachers to live off the generosity of those taught, this precept could aptly be used to encourage those taught to be generous.

3. Paul fails in 1 Corinthians 15 to refer to any dominical teaching on the resurrection from the dead, but he does not so fail in 1 Thessalonians 4.15-17:

For this we say to you by the word of the Lord, that we who are alive and remain to the advent of the Lord will not precede those who have fallen asleep, since the Lord himself, with a shout, with the voice of the archangel and with the trumpet of God, will descend from heaven, and the dead in Christ will rise first. Afterward we who are alive and remain will be raptured up together with them in the clouds to meet the Lord in the air, and thus we will be with the Lord for all time.

Paul knew a dominical word on the resurrection of the saints* after all. The argument from silence, as applied to 1 Corinthians 15, would have led us astray on this point.

* 1 Thessalonians 4.15-17 does not exactly match any dominical saying from the gospel traditions, though it resembles parts of the Olivet discourse (which, however, does not explicitly mention resurrection). But the point remains that Paul had a word from the Lord pertaining to the resurrection, yet failed to use it in 1 Corinthians 15.

In each of these cases Paul the apostle (A) was in a position to know a dominical saying relevant to his subject matter (and, as it turns out, did in fact know one) and (B) had good reason to use that dominical saying, if we consider the relevance of a command of the Lord to be a good reason to use that command.

Each of these cases, then, passes the first condition but fails the second. I might add, however, that the only reason we know for certain that each case passes the first condition is because Paul actually did use the dominical saying in question elsewhere in his extant epistles. Had he never mentioned the dominical words on divorce, on living from the gospel, or on the resurrection, how would we know for certain that he knew those sayings, even if they did exist? We cannot necessarily expect him to know everything the Lord ever uttered. As for the second condition, relevance to the subject matter is apparently not always a good enough reason to suppose that Paul would have referred to the saying had he known it.

One inconsequential Tertullianic silence.

Let me offer, finally, one last illustration of the danger of the argument from silence. I have not read Earl Doherty, The Jesus Puzzle, so cannot comment on his use or nonuse of the argument from silence in that book, but he does occasionally drop by the IIDB, and I have read a few of his posts there.

In one of his posts (June 25, 2006) Doherty quoted Bart Ehrman, The Orthodox Corruption of Scripture, as follows:

The verse [Galatians 4.4] was used by the orthodox to oppose the Gnostic claim that Christ came through Mary "as water through a pipe," taking nothing of its conduit into itself; for here the apostle states that Christ was "made from a woman" (so Irenaeus... and Tertullian...). It should strike us as odd that Tertullian never quotes the verse against Marcion, despite his lengthy demonstration that Christ was actually "born." This can scarcely be attributed to oversight, and so is more likely due to the circumstance that the generally received Latin text of the verse does not speak of Christís birth per se, but of his "having been made" (factum ex muliere).

Ehrman has here used his own argument from silence. Tertullian never cites Galatians 4.4 against Marcion, even when Galatians 4.4 would have supported his point that Jesus had really been born. Ehrman takes this silence to mean that the Latin text of Galatians 4.4 which Tertullian used had, like many of our extant manuscripts, the word made instead of born.

But Doherty wished to press this argument from silence a bit further. He wished to argue that the Latin text of Galatians 4.4 used by Tertullian lacked the entire phrase made from a woman altogether:

Iíll remark here that Ehrmanís reasoning is a little off. If Irenaeus and Tertullian could use genomenon (come, or "made from") in Greek, why could Tertullian not use the Latin equivalent, especially since even this version would have been useful? Wouldnít another explanation recommend itself: that Tertullianís version contained neither phrase, indicating that his Old Latin text was derived from an earlier version of Galatians which did not have "come/born of woman"?

It is true that Irenaeus finds the phrase made from a woman (instead of born from a woman) in Galatians 4.4 useful in its own right (refer to Against Heresies 3.16.3; 3.22.1). So Doherty argues that Tertullian would have found it useful, too; his failure to use the Latin version of this phrase against Marcion indicates that his Latin text of Galatians 4.4 lacked the phrase altogether. To fill out our two conditions, Doherty argues (A) that Tertullian would have had either factum ex muliere or natum ex muliere in his Latin manuscripts, if the phrase was extant in his time, and (B) that he surely would have used that phrase if he had it.

But there is a problem. Tertullian does use the phrase factum ex muliere, but not against Marcion. He uses it in On the Flesh of Christ 20:

Sed et Paulus grammaticis istis silentium imponit: Misit, inquit, deus filium suum, factum ex muliere.

But even Paul imposes silence upon these grammarians.* He says: God sent his son, made from a woman.

* Tertullian calls his opponents grammarians here because the context is an argument over prepositions.

With that, the second condition (that Tertullian would have used the Latin phrase against Marcion had he known about it) falls to the ground. It turns out that Bart Ehrman himself knew, as we might expect, of this line in On the Flesh of Christ 20. Doherty, in quoting Ehrman, had blocked out (for whatever reason) the reference to On the Flesh of Christ 20 with an ellipsis:

...for here the apostle states that Christ was "made from a woman" (so Irenaeus... and Tertullian...).

So, while Ehrman was certainly employing an argument from silence (Tertullian failed to use Galatians 4.4 against Marcion) to show that Tertullian did not know the natum ex muliere variant, he was merely using that argument from silence to support a much more substantial argument, the argument from On the Flesh of Christ 20, in which Tertullian actually uses the factum ex muliere variant.

This is, in fact, how I recommend using the argument from silence, to wit, as support only. It can only rarely stand by itself. The historian leans upon it at his own peril.