Introduction
One of the oldest philosophical questions within Judaism (and outside of it) is the possible contradiction between Hashem's Omniscience (knowledge of all) and our own free will. The question was perhaps made famous by the Rambam, but was already recorded many years previously.
The apparent contradiction is that if Hashem knows everything, including future events, a human (or any other creation) cannot possibly choose to act differently to the way Hashem knows he will act. Thus, even though a person is unaware of what he himself will decide in the future, he cannot be said to be acting out of free will.
Whenever there are two seemingly contradicting principles, there can only be three options:
2) The first principle is wrong.
3) The second principle is wrong.
It may come as a surprise to many, but to some extent, all three possibilities have their advocates among the rishonim. We will start by analysing the first option, which is perhaps the most commonly accepted and the position of the oldest Jewish written works.
Is there even a problem?
Although many do believe that this contradiction can be resolved, there is considerable variance within this school of thought as to how easy the resolution is. R' Sa'adia Ga'on (10th century, Egypt/ Eretz Yisrael / Baghdad) dismisses the whole issue as a non-question – although Hashem knows what a person will do, this knowledge is not the cause of the person's choice. On the contrary, the person's choice affects Hashem's knowledge in the past.[1]
What was obvious to R' Sa'adia was far from obvious to others. The idea that the future can affect the past (Retrocausality) is a topic debated by philosophers and is not limited to theological discussions.[2] It is also worth noting that R' Sa'adia's argument is not dependent on any Divine qualities that are beyond our understanding – according to his logic, there would also be no inherent problem in a human foreseeing a free will choice made by another human with 100% accuracy.[3]
By contrast, R' Bachaye ibn Pekuda (11th century, Spain) writes that were we able to understand the resolution of this contradiction, Hashem would have revealed it to us.[4] In other words, although there is a resolution, we are incapable of understanding it.
This view is echoed by the Rambam, although he adds more of an explanation as to why we are incapable of understanding. While humans have characteristics that are separable from their essence, Hashem does not. No positive characteristics can be attributed to Hashem – He and 'His knowledge' are One. As this is a concept that us humans cannot fully fathom, we are also incapable of understanding how Hashem attains knowledge of human choices. The only thing we can be assured of is that we do indeed have free will.[5]
The problem with this is that even after we accept that Hashem's capability of seeing the future is not something we can understand, we are left with a contradiction from the human point of view. Unless we accept the retrocausality of R' Sa'adia Ga'on,[6] Hashem's unerring knowledge of the future means that there is no possibility that we will in fact decide something different.
It would therefore appear that when the Rambam insists that we have free will, this only means that our decisions are made willingly. We do not actually have any real option to do something different to that which Hashem knows we will do.[7] This is notwithstanding the fact that the Rambam spends this entire chapter stressing that free will is a fundamental principle of the Torah[8] and that those who deny it are foolish.
If this understanding of the Rambam is correct, we can better understand the criticism of the Ra'avad of his words. The Ra'avad writes that a wise person should not raise difficult questions without resolving them, lest this causes any heretical thoughts in the mind of the reader. This criticism is more understandable if the Rambam himself could have helped us by clarifying that free will is more limited than we may have thought.[9]
'Inability' to do the impossible
We now move on to the second option above. Suggesting any limits to Hashem's knowledge may seem heretical, but this is based on a lack of both knowledge and understanding. Nowhere do Chazal state in an unequivocal manner that Hashem foresees all future events (although it is quite clear from Tanach that He does foresee many of them).[10]
The Rambam's himself, while listing Hashem's Omniscience as the tenth principle of faith, does not mention the future. His statement of Hashem's knowledge of all future events only appears in the context of his discussion of free will. It appears that the Rambam derived this concept from the general principle that Hashem is perfect and is lacking nothing.
However, as the Rambam writes elsewhere, it does not make sense to describe Hashem as able to do the logically impossible, such as creating another entity like Him.[11] It follows that if knowledge of our future decisions is a direct contradiction to our free will, Hashem's 'lack of ability' to create such a contradiction would not reflect any real lack. He simply chose to grant us free will, thus 'forgoing' His own prior knowledge of what we would choose.
This is the position of Ralbag (1288-1344, Southern France).[12] The Rambam could not follow this line of logic, as his view is that time itself is merely a creation of Hashem.[13] The idea that Hashem's own creation could limit Him is itself a contradiction.
One thing should be clear. As the view of the Ralbag does not attribute any lack of perfection to Hashem, and the dispute revolves around the philosophical understanding of the concept of time – this view is well within the acceptable limits of Jewish religious belief.[14]
Determinism
Perhaps even more surprisingly, R' Chisdai Kreskas goes to the opposite extreme. Preferring to accept the position that Hashem does foresee all future events, as well as due to other philosophical arguments, he concludes that human actions are always just the net results of all the natural causes acting on a person (both from within the person's nature and from external sources). Although we do not feel anything forcing us to act in a certain way, the reality is that we have no choice.
This leads to the obvious question as to the justification of reward and punishment for our actions, which are not in our control.[15] R' Chisdai is clearly aware of the problem and writes at great length to attempt to resolve this. He first claims that reward and punishment are necessary tools with which to influence our decisions for the good (these are some of the many factors that force us to act in the way we do), and this end justifies even the punishment of the innocent.
However, he goes on to ask that if this is the case, why can't there be punishment even for those who act entirely under noticeable duress? His conclusion is that in fact reward and punishment is not for the actions themselves that a person carries out, rather for the frame of mind in which they act. One who happily serves Hashem earns reward for his joy, and the same is true for the converse.
With this, R' Chisdai Kreskas concedes that regarding our feelings, we do in fact have free will. Although he does not revisit the question of Hashem's prior knowledge of everything, the clear implication is that Hashem can only foresee what our actions will be, not our mental attitude towards them. Thus this view actually combines both the second and third options, limiting both Hashem's knowledge of the future and our free will!
On a philosophical level, it is hard to see what is gained by this strange combination. Nevertheless, as R' Chisdai Kreskas does not deny the concept of reward and punishment or the justness of the ways of Hashem, this view also certainly remains within the limits of legitimacy.
May we use our free will (whatever level we have) for the good, and on Yom Kippur attain forgiveness for any bad decisions we may have made in the past!
[1] Emunos v'Deyos, Ma'amar 4.
[2] One of the main arguments against the possibility of retrocausality
is that were it possible to have an effect on the past, this would lead to time
paradoxes. On the other hand, retrocausality is observed within quantum
mechanics. Therefore, R'
Michael Avraham argues that retrocausality is only possible within deterministic
processes and not cannot be affected by decisions made with free will.
[3] Of course, humans cannot do anything with 100% accuracy. However,
this is due to human qualitative deficiency, rather than any logical
impossibility.
[4] Chovos Halvavos, Sha'ar 3 (Sha'ar Avodas HaElokim), chapter 8.
[5] Hilchos Teshuva 5:5
[6] The Rambam appears not to – otherwise the question would not
require an answer unfathomable to us.
[7] Thus the Rambam is not as strong an adherent to option 1 above –
his position is somewhere between option 1 and option 3.
[8] R' Nahum Rabinovitch calls this the 'fourteenth principle of faith'
of the Rambam (Mesilot Bilvavam, p.54).
[9] I understand the answer that the Ra'avad suggests, that Hashem
knows 'like the knowledge of astrologers, who know through an external power
how another will act', as being equivalent to my understanding of the Rambam.
The Ra'avad's conclusion, that this answer is not enough, is presumably because
of a reluctance to accept that free will is limited in this way.
[10] The Rambam understands that a statement to this effect, together
with the principle of free will, appears in the Mishna in Pirkei Avos (3:15).
However, this explanation is not obvious from the context of the Mishna, and
neither is it unanimously agreed upon (see for example Bartenura for an
alternative explanation).
[11] Moreh Nevuchim 3:15
[12] Milchamos Hashem, Ma'amar 3, Chapter 4. This was also later the
position of the Or HaChayim (Bereishis 6:5) and the Sh'lah (in his
introduction).
[13] Moreh Nevuchim 2:13. It is however unclear how the Rambam knew that
time was a creation.
[14] See The
Limits of Legitimate Belief, where we supported the view of the Ra'avad,
that the same is true even regarding the mistaken view that Hashem has a
physical form.
[15] Or Hashem, Ma'amar 2, Klal 5.