Causality: E. Praying for What Has Already Happened and Praying for the Past (Column 464)

Introduction

In the previous column I discussed the halakhic prohibition to pray and ask about what has already occurred, and essentially about a miracle in general. We saw there the difficulty in this prohibition—namely, the very distinction between praying for a miracle and a permitted prayer. As I explained there, every prayer is a request for a miracle.

In the background, ostensibly, stands the assumption that the past cannot be changed. Yet in the end we saw that that discussion was not really about changing the past but about requests regarding the present. Here I wish to address the question of changing the past itself. Before that, however, I will sharpen the difference from the previous column.

Between praying for what has already happened and praying for the past

In the previous column we saw that even when Hazal and the halakhic authorities speak of “praying for what has already happened,” the intention is to prayers about the present and future—except that this is a future already fixed by the state of affairs in the past and present. Thus, for example, we saw the case of requesting the sex of a newborn. I pray that the fetus in my wife’s womb will be male, but for me it suffices that he be born male; I don’t care what the sex was in utero up to the birth. For all I care, God need not change the past but can change the fetus’s sex a moment before birth (perform sex reassignment in utero). And yet, the Talmud says that since being male is determined by the state at the beginning of pregnancy, from forty days onward one may no longer pray for that. This is what, in halakhic jargon, is called a prayer for what has already happened, even though—as noted—it deals with the present and future. Likewise, when a person who sees a fire in the city asks that the victims not be his household, he is not really praying about the past but about the present. He wants his household to be alive now. If they died in the fire, as far as he is concerned God can revive them now, and that would be perfectly fine. That is, even there the request is about the present, not the past.

At first glance, praying literally for the past is futile and meaningless. Why should I care about something that happened in the past? Requests always concern the present and future. And yet, at least theoretically, there is room to discuss a request about the past itself. Think, for example, of a person who sees the fire but does not only want his family to be with him now—he wants them not to have suffered in the fire at all. In that case he is asking about the past itself (that they not be the ones who burned) and not only about the present. What shall we say about such a request?

This is already stranger and more far-reaching than what we saw in the previous column. It is a prayer about an event that itself was in the past, not only about its implications for the present and future. Thus, for example, I ask whether one may pray that Nelson win the Battle of Trafalgar (which did in fact happen in 1805, where the legendary British admiral Horatio Nelson defeated Napoleon’s fleet, at the cost of his own life). Alternatively, may I pray that Napoleon win the Battle of Waterloo (which did not happen; in 1815 Napoleon was defeated by the coalition of the Prussian von Blücher and the British Duke of Wellington)? There are two kinds of prayer here: I ask that something that happened turn out otherwise, or I ask that something that happened indeed be as it was.

At first glance this is a meaningless string of words: mai de-hava hava—what has been, has been. What does it mean to pray about something that has already been?! Note that this is not a question about God’s abilities—whether He can or cannot change the past. The past already was. The “change” is not even conceptually definable. What would it mean to change the past—erase it? Cause it never to have occurred? But it did occur. Am I asking that, although it occurred, God decree that it did not occur? What sort of request is that? This is a problem of definition, not merely of physical ability. One might perhaps say that we are speaking of changing our consciousness—that is, the petitioner wants God to cause us to live from now on with the sense that it did not happen. But that is a change having no connection to the past, only a change in present consciousness: a request to flip a few bits in everyone’s memory. That, however, is not a prayer about the past, but a prayer about what has already happened—the topic of the previous column.

To sharpen this further, I will add that, on this view, it is even stranger to pray for Nelson’s victory (which did in fact occur) than to pray for Napoleon’s victory (which did not). Regarding Napoleon—if we can even speak of God’s ability to change the past—then perhaps there is point in praying for his victory so that the result be altered retroactively (again, assuming this is definable). I am dissatisfied with what happened and ask God to change it. But regarding Nelson, if he already won, what is the point of praying for it? What exactly do we expect God to do here? In light of the odd assumption that God can change the past (and thus it is possible to pray for Napoleon’s victory), perhaps one could say we may pray that God not change Nelson’s victory. But the whole business seems decidedly odd.

Astonishingly, there are philosophers and theologians who argue that there is indeed point in praying for both kinds of cases (to change or to preserve what happened). For example, Michael Dummett of Stanford wrote an article entitled “Bringing about the past,” in which he forcefully argues that indeed there is. Tim Mawson of Oxford says likewise in his article, “Praying for known outcomes.” Incidentally, Mawson ends his article with an impassioned call to all of us—loyal Britons that we are—to join him in praying for the Duke of Wellington’s victory at Waterloo (from 1815):

Since you value Wellington’s victory as I do, I end with an appeal to you—if you have no more pressing matter to which you must attend upon finishing this article and before turning your attention to your next activity—to join me in a brief prayer for the success of the Duke of Wellington in what we have every reason to believe was, in Wellington’s words, “the nearest run thing you ever saw in your life.”[1] We also, of course, have every reason to expect our prayers to be answered in the way we should like, since they have already been answered so impressively—Wellington did indeed win—and therefore we must also thank God for that. Amen.

Blessed is He who speaks and does. Blessed is He who hears the prayers of His people Israel with compassion; amen and amen.

These fellows emphatically advocate prayers to preserve an outcome already achieved in the past—and, of course, prayers to change outcomes that did not occur. They do, however, phrase the claim somewhat cautiously: according to the standard theistic model, such prayers have meaning. Perhaps the intent is to critique the theistic model rather than to state a thesis they are in fact prepared to defend. Still, if eminent philosophers from Oxford and Stanford think one can present such a position, it is at least worth examining, no?

C. S. Lewis’s claim

Our acquaintance C. S. Lewis—the author of Narnia and, as I like to joke, of “Shas”—has already been favorably mentioned on this site. Beyond being a successful and celebrated children’s author, he is also one of the clearest and sharpest philosophers I know, and he was gifted with a keen sensitivity to problems that only began to bud in his day (he died in 1963). Among other things, Lewis also put his hand to the issues of miracles and praying for miracles, and he even wrote a book devoted to these topics.[2] Mawson opens his article with a passage from that book of Lewis’s:

When we pray about the outcome of, say, a battle that has already been fought, and suppose that its outcome is unknown to us… the thought will commonly occur to us that (if only we knew it) the matter is already settled one way or the other. I do not think this a good reason for ceasing to pray for it. The event is indeed settled… but one of the things taken into account in the divine decision to give victory to the side that won—one of the things which really brought it about—may be that prayer which I am offering now. Therefore, paradoxical as it may sound, I conclude that we can at noon become part of the causes of an event happening at ten o’clock.

It may be asked, “If I stop praying, can God go back and change what has already happened?” No. The event has already happened; and one of its causes was that you ask such questions instead of praying. It may be asked, “If I begin praying, can God go back and change what has already happened?” No. The event has already happened; and one of its causes is your present prayer. Thus the event’s outcome genuinely depends on my choice…

Lewis’s discussion proceeds on the assumption that even God cannot change the past. He, however, assumes that God stands above the time axis, and therefore for Him there is no problem “seeing” the future as the past (even if not changing the past). The entire timeline and the events occurring along it lie spread out before Him; among other things, He then sees what I am doing now. But, claims Lewis, this was visible and known before Him a thousand or ten thousand years ago as well. If so, in 1805, when deciding who would win at Trafalgar, nothing prevents Him from taking into account the fact that I am praying for it now (some 220 years later), and this, precisely, constitutes part of the considerations in His decision to give the victory to Nelson (or to deny it to Napoleon).

And what of praying that Napoleon win at Trafalgar (or at Waterloo)? In his view, there is no bar to praying fervently for that as well—except that in that case, alas, we already know that our prayer was not answered. But hey, don’t forget that no prayer or tear of a righteous person goes unanswered. You, believers the children of believers, will surely agree that prayers for the future are not always answered (note the understatement of the century); so too there is no bar to praying for the past (and not being answered).

In short, there is no difference between praying before an event or after it, for God knows the future as the past, and He takes the prayer into account when deciding the outcomes of the battle or event. Sometimes He will heed the prayer and sometimes He will not—but that is true both of prayers about the future and prayers about the past. If so, why should we not pray about events in the past?! Incidentally, given my view that God is not involved in the world and does not respond to prayers, I can only bolster his words: one who prays about a future event even though he knows it will not help—why should he not pray about the past? Granted, that too will not help, but futility has never stopped us all from offering prayers and making requests with great devotion.

Mawson’s discussion

In his article, Mawson wears us out with a long and tedious analysis that adds little beyond Lewis’s argument. He merely proposes options that blunt the edge of his revolutionary suggestion. For example, regarding prayer about the future, he says there is a possibility that what is about to occur will in fact not occur. Think of Wellington seeing plumes of dust from a large military force approaching Waterloo. He wonders whether it is Blücher the Prussian, his ally, or perhaps another force allied with Napoleon—and on that the battle’s outcome depends (indeed, that is how it actually was: the battle hinged on who arrived first, and it was unclear until the last moment who had come). When the dust cloud disperses and Wellington sees Blücher’s flag—his ally—he understands this means they will win the battle. Amid his tears of joy and excitement, Mawson asks whether, in such a case, there is still point in asking and praying about the battle’s outcome.

His answer is affirmative. But within his words he explains it by saying that it is always possible that the outcome will be otherwise, albeit with very slim probability.[3] In doing so he of course empties Lewis’s claim of content. Such a prayer is no novelty; it is simply a regular prayer about the future. As long as the possibility that Wellington will lose exists, the relevance and meaning of his prayer are obvious (recall that the discussion here proceeds under the assumption that God is indeed involved in the world).

So much for praying about the future. As for praying about the past, Mawson performs a similar maneuver. He argues that although we today know that Wellington won, it is possible that the historians erred and presently new information will emerge showing that he actually lost. If so, there is still point in praying about it, since what blocks prayer for the past is our knowledge of what happened. That might be psychologically difficult, but there is no essential bar. This is, of course, nonsense, because the futility of praying for the past has nothing to do with whether I do or do not know what happened. Even if I do not know what happened, still whatever happened happened and cannot be changed from the present backward (as noted, that was also Lewis’s assumption). Therefore there is no point or value in praying about it.

Let me sharpen this via one of Mawson’s own arguments. He notes, in passing, that it was extremely important to the world (not only to the British) that Wellington won at Waterloo. Had the result been otherwise, the world would have been much worse. Therefore, he argues, it is certainly important to pray for it. Now let us suppose, for the sake of argument, that the historians were wrong and in fact Napoleon was the victor. Under that assumption, the reality familiar to us today is the result of Napoleon’s victory. It then turns out that this victory was not as terrible as he assumed, and the need for Wellington’s victory is not there. If so, by his own lights there is no need for this prayer, since the world is as it is and the events in the past that produced it (whatever their nature) have already done so. This is, of course, an anecdotal and unimportant argument, but it highlights the inconsistency embedded in Mawson’s picture.

My claim against his position is that the futility of praying about the past is not connected to my knowledge of the past, but to the fact that it has already occurred. Even God Himself “cannot” change the fact that it has occurred (for the meaning of “cannot” here, see, for example, column 302, in the thread here, and many more). Whether we assume that Wellington won or lost—and even if I do not know—what’s done cannot be undone: mai de-hava hava. I already noted that this is easier to illustrate with a prayer in favor of Wellington than with a prayer against him. Think what the purpose of such a prayer would be. Wellington already won. Am I asking God to change anything? I am perfectly content with the status quo.

Back to Lewis

If we return to Lewis, he argues that although there is no meaning in speaking of changing the past, nevertheless the victory was achieved in part thanks to my present prayer; therefore there is point in praying. God is above the time axis and so can take the future into account as He does the past. I will grant this for the sake of argument, and still I do not understand the claim. Even if I do not pray, the victory has already been achieved in the past—only, on his view, if I did not pray, it occurred not thanks to my prayer but for other reasons. In any case, the victory was decided in the past whether I pray for it or not. So why pray?! Should I pray so that I have “a share” in an event that has already occurred—so that Wellington’s victory be due, in part, to me? It is commonly thought that the purpose of prayer is to obtain some result (this reminds me of the marvelous thesis of “hishtadlut”—see column 279). This seems to me akin to offering an impassioned prayer that either Wellington win or lose or that it be a tie (incidentally, this prayer is dedicated to the blessed memory of Zermelo, of blessed memory).

Note that my claim is not against God’s ability to foresee the future and even take it into account in making decisions in the present,[4] nor against God’s very involvement in the world (i.e., the assumption that He is the one who decides who wins a given battle). Here I argue only against my motivation to pray about the past. A prayer about the past is meaningless even if God could, in principle, have taken it into account.[5]

In short, despite my deep solidarity with the British people and its values, it will, alas, be hard for me to heed Mawson’s impassioned call and pray for Wellington’s victory. I shall console myself with the certain knowledge that had I prayed, my prayer would have been answered. It is no small thing to pray knowing your prayer is answered. Surely God would also have placed my salty tears in His flask.

Fatalism

So much for Mawson. Now to Dummett, who in his article compares prayer for the past to fatalism. His claim is that just as we do not accept fatalism—which negates prayers regarding the future—so too we should reject the arguments against praying for the past.

A fatalistic position holds that everything that happens happens necessarily. We are delivered into the hands of fate, and it is forced upon us. What I have described so far, however, is determinism. Fatalism is a psychological and perhaps axiological layer built atop determinism. Typically, the fatalistic layer contains two components: (1) a psychological state—a passive stance of acquiescence in reality and fate; and (2) a value-dimension—attributing value to that passivity and acquiescence. In effect, fatalism is a kind of exalted indifference (moralists call it “hishtavut”).[6] The Druze say that every bullet has an address; therefore if someone fell in battle or was killed in an accident, it was bound to happen anyway. That is the deterministic component. One conclusion that can be drawn from that deterministic picture is that there is no point in grieving, and no point in praying to change the future that awaits us—and here we enter the realm of fatalism.

At the beginning of his novel Appointment in Samarra, John O’Hara brings the following Muslim (Sufi) tale:

A merchant in Baghdad sent his servant to the market to buy provisions. In a little while the servant came back, white and trembling. “Master,” he said, “just now in the marketplace I was jostled by a woman in the crowd, and when I turned I saw that it was Death that jostled me. She looked at me and made a threatening gesture. Please, lend me your horse, and I will ride away from this city and avoid my fate. I will go to Samarra and there Death will not find me.” The merchant lent him his horse, and the servant mounted it and as fast as he could he rode away. Later that evening the merchant went down to the marketplace and saw Death standing in the crowd, and he asked her, “Why did you make a threatening gesture to my servant when you saw him this morning?” “That was not a threatening gesture,” said Death. “It was only a start of surprise. I was astonished to see him in Baghdad, for I had an appointment with him tonight in Samarra.”

I assume there is no need to note the influence of mythic Oedipus on this outlook.

The comparison to fatalism

Dummett presents the fatalistic argument as follows (at the bottom of p. 345). He begins with a tautology: either you are going to die in this raid or you are not. For everyone’s sake I shall allow myself to ignore quibbles about logic and the law of the excluded middle; they are all fairly silly.

The next two steps of the argument run in parallel:

1a. “If you are going to die in this raid, you will die regardless of the precautions you take.”

1b. “If you are not going to die in this raid, you will not be killed even if you take no precautions at all.”

These are two claims of the type: “If p, then for every q, if q then p.”

The next step is to draw conclusions from those claims:

2a. If you are going to die in the raid, then any precaution you take is ineffective.

2b. If you are not going to die in the raid, then any precaution you take is superfluous.

Dummett engages in casuistry about the meaning of the “if” in these sentences, but allow me to be brief and indicate the error in the fatalist’s argument: true, I am going to die in battle—but that is caused precisely by my not having taken precautions. Had I taken them, I would not have died. In fact I did not take them, and therefore I am indeed going to die. This actually brings us back to the argument about logical determinism (see column 459), where I noted logic’s indifference to the time axis. If in fact Reuven dies in the raid, then indeed even the day before it was true to say that tomorrow he is going to die (though then we did not yet know it). But from the fact that tomorrow Reuven will die one cannot deduce that he must die, and therefore one cannot deduce that he will die regardless of the steps he takes today. On the logical plane, if indeed he will die tomorrow, then the statement that he will die tomorrow is true even today. But it is not true that this fact is already fixed today. Thus the fatalistic argument fails on the logical level.

Of course the “precaution” in question here can also be prayer. The fatalist’s conclusion is that there is no point in praying; but he is wrong. On the logical level, the outcome achieved will be based in part on the prayer. In the continuation of his remarks, Dummett builds his argument on the comparison between prayer about the past and prayer about the future. His claim is that the same flaw present in the fatalist’s argument regarding prayer about the future exists also with respect to prayer about the past. And just as the conclusion is that there is point in praying about the future, so too there is point in praying about the past (very much like Lewis’s argument above).

The structure of the argument that negates prayer about the past is as follows. If Wellington lost the battle, then prayer is ineffective; and if he won the battle, then praying is redundant. This precisely parallels 2a and 2b regarding the future. And if the logical structure of those claims is faulty, then so too the claims about the past are faulty. Note that the flaw I identified in the fatalistic argument is a logical flaw unrelated to the time axis; therefore this flaw should also obtain regarding claims about the past. The same logical structure suffers from the same flaw. His conclusion is that there is point in praying about the past as there is about the future.

I will not prolong myself with his casuistry, for there is a basic logical error in his foundational analysis. Indeed, the logical argument against praying for the past fails—exactly as with the future. But that only shows that there is no logical problem in praying for the past. Does that necessarily mean there is no other problem with praying for the past—one that does not exist with regard to praying for the future? Absolutely not. As we saw above, the fundamental problem with praying for the past is not logical but pertains to the time axis. If the event has already occurred, then there is no point in praying about it, because it has already happened (whether I wish to change it or to preserve it). That is not a logical problem but a factual one (the inability to affect the past), or even definitional (what does it mean to change the past now? mai de-hava hava!). The time axis is, of course, asymmetric between past and future, and therefore the fact that this problem exists for prayer about the past does not entail that it exists for prayer about the future.

Conclusion

The conclusion is that unlike a “cry about what has already happened,” in Talmudic parlance—about which there is a prohibition to pray (but no logical bar)—a prayer about the past itself is not a prohibition but a definitional problem. Not only can one not undo what has been done, and God “cannot” change the past; the prayer itself is undefined. In principle one may pray to change the present—that is, to change all our consciousness regarding the Battle of Trafalgar or Waterloo—but I explained that this is not a prayer about the past (though it is likely included in the prohibition discussed in the previous column). A prayer to preserve the past does not even request any change (neither in the past nor in the present), but simply to leave things as they are. About that there may not be a prohibition to pray, but it sounds rather silly.

Note that this analysis sharpens the meaning of the temporal component of the causal relation (see column 459). At the beginning of the series I noted that a cause cannot change an event that occurred before it. That is the meaning of the temporal component of causality. Here we clarified why this is so: one cannot change the past because it has already happened, and therefore one cannot cause or bring about the past. Thus a cause at a given time cannot affect an effect at an earlier time. My point is that the problem with praying about the past mirrors the problem of causal influence on the past. The problem is definitional, bound up with the nature of the time axis. It is so severe that it would seem that even God Himself cannot.

In the next column I will continue from this point and examine the relation between praying for the past and traveling back to the past, and various kinds of impacts on the past in the legal sphere.

A brief note on philosophy and philosophers

A final remark on the two articles discussed here. Time and again I am surprised at the nonsense that respectable scholars of philosophy, even those working at prestigious universities, pour upon us in their papers. I am no less surprised that such nonsense passes peer review. No wonder the field has such a bad reputation—undeservedly. The problem is not in philosophy but in philosophers, or in researchers of philosophy.

That said, it is important to note that to claim something is nonsense, an initial impression of empty pilpul does not suffice. Such a feeling about a philosophical discussion crops up often, and sometimes unjustly. For example, in the first notebook I mentioned the critique by Dawkins and many others of the Ontological Argument, and I showed there that while the feeling may be present, the critique stems from a lack of philosophical skill. They were satisfied with a sense of nonsense without engaging in a substantive critique of the argument—one most of them likely did not know how to conduct. Substantive critique requires analysis and pointing out the flaws in the arguments. That is what I have tried to do here. After one points out the flaws, one may add that it is nonsense. The initial feeling can guide us and set a direction, but it does not replace substantive analysis.

[1] The translation is mine, and there is a phrase that was not entirely clear to me: “the nearest run thing you ever saw in your life.” I hope I rendered it reasonably.

[2] Miracles, A Preliminary Study, C. S. Lewis.

[3] He nonetheless insists that one can still say that Wellington knows he has won the battle, because “knowledge” is a concept usable even where there is no certainty. I fail to see the interest of that semantic remark. It reminds me of my critique of David Enoch’s discussion of statistics in law, where he treats the term “knowledge” similarly in the legal context (see column 228).

[4] I do not agree with that either, as I explained at length in the series on foreknowledge and free will (299–303). But I do not need that premise here.

[5] Toward the end of column 299 I explained the difference between the question of how God knows the future and another question: granting that He knows, can we still say that I have free will? I showed there that these are two different questions. The distinction I have made here is quite similar. One might argue that in the nineteenth century God knows that I will pray two hundred years later. But that does not necessarily mean there is any point for me in praying. This is also akin to the argument I advanced in that series via Newcomb’s thought experiment.

[6] The source of this value is likely Buddhist. See, for example, here.