Machine Will Do and Machine Will Learn

by Rabbi Aryeh Klapper

Dear Friends,

Thank you so much for reading our Torah. Should you be so moved, here is the link to donate to CMTL. Your support is deeply appreciated and utterly necessary for our work to continue and our Torah to spread. G-d willing we’ll be in touch with more detail about that work early next year.

In gratitude,

Aryeh Klapper

Dear Rabbi Klapper,

Many years ago, I asked you whether the positronic robots in Isaac Asimov’s I, Robot were obligated to perform mitzvot. I think you answered with sources about golems and orangutans; all I remember is being so happy that you took the question seriously. 

Now I have a hopefully more grown-up version of the same question: What are the responsibilities of programmers and users for the decisions and actions of artificial intelligences? Toward artificial intelligences? Do artificial intelligences have religious, ethical, or moral responsibilities?

I look forward very much to your reply.

With all best wishes,

Yoni Kohenson

Dear Yoni:

So great to hear from you! I remember the question well – I was 18 and you were 12, and it was the first time anyone had asked me a halakhic question with some sense that my answer would have authority. Happily my father z”l had taken me a few years earlier to a shiur on AI at an AOJS convention, and I remembered the speaker quoting Tif’eret Yisrael on the personhood of orangutans[1], and the amazing machloket acharonim about whether golems count to a minyan. (Q. Why did Rabbi Eliezer need to free his slave to make a minyan, when he could just have made a golem? Golems must not count! A. It would have been yuhara = arrogant/hubristic to make a golem for that purpose.) But those discussions assume that the question is whether other beings can be assimilated to the category of human. 

Another useful starting point – I’m not sure I had read it then – is Rabbi Norman Lamm z”l’s “The Religious Implications of Extraterrestrial Life”. Rabbi Lamm works hard to show that Jews should not be frightened by the possibility that the universe contains other types of persons. What makes humans valuable is our quiddity/whatness and not our scarceness or uniqueness (although our value as individuals may be related to our ultimate uniqueness).

I’ve put a lot of weight already on the word “persons”. So let’s try to define it. I propose that it refers to the presence of three elements: 

1. consciousness – by which I mean awareness of oneself as an integrated being. I’m consciously ignoring Sartre’s argument one cannot know that the cognizing and cognized selves are the same.

2. prudential judgment – by which I mean the ability to know which means are how likely to produce which consequences. I understand that human judgment is highly imperfect in this regard. 

3. free will – by which I mean “hard” free will, the ability to make choices not determined in advance either by one’s environment or one’s own being. I am aware that some Jewish thinkers probably deny that human beings have this sort of free will, and that its existence cannot be philosophically demonstrated. 

Requiring all three components of personhood for individuals creates grave moral difficulties with regard to infants (especially hypoencephalic infants), the insane, and so forth. The usual response is to require them for the species, and then extend the species’ umbrella to individuals. Extending personhood to nonbiological beings would problematize that – do computers have species? It might drive us to use the language of essences/souls instead.

Rabbi Lamm points out that halakhah is Terracentric. Earthlings finding themselves on a planet with permanent daylight everywhere, or habitable only at the poles, might adapt Talmudic models of time like “the person lost in the desert” and so forth. But it would make no sense for the Torah we have to be revealed on a planet without sunsets and pigs and other things that the Torah specifically regulates. G-d would give a different Torah to ETs who met our criteria for personhood. The same logic applies to persons who lack biological bodies, because “The Torah was not given to the ministering angels”.

This same issue applies applies to at least the prohibitions against incest and eating limbs from live animals among the Seven Noahide Commandments. So I don’t think that one can assimilate AI persons to the category of Noahides either, nor do I think ETs are Gentiles.  

The best model I have for nonhuman persons is Rav Asher Weiss’s position on publicly held corporations. Rabbi Weiss holds (cf. the U.S. Supreme Court in Citizens United) that corporations are halakhic persons who are bound neither by Jewish halakhah nor Noahide law, and toward whom Jews owe no halakhic duties. Therefore publicly held corporations, even if Jews own most of their stock, can both charge and be charged interest to each other and to individual Jews. Rabbi Weiss nonetheless writes (Minchat Asher 1:105) that 

“it seems correct that in all matters related to the prohibition of theft and robbery and the like, that are rational mitzvot between humans and their fellows – it is certainly obvious that these mitzvot are obligatory even on a corporation. It is forbidden to steal and it is forbidden to steal from it, and it is obvious that this body which has within it free-willed decisionmakers must behave in the manners of justice and integrity, because the world stands on truth and on law and on peace… but everything related to Torah prohibitions that are nothing but decrees of Scripture, such as chametz on Pesach, Shabbos, interest, and the like – there is fundamentally no prohibition of these with regard to the money of the corporation”. 

I think the same is likely true of AI persons. 

Rabbi Weiss gives no clear mechanism for constructing this reason-based system. I think AIs would have to play a significant role in recognizing and where necessary constructing any system under which they could be held accountable. For example, I am not at all certain that biological beings are likely on their own to properly determine how the concepts of life and death should be applied to AIs.

Corporations are not in any way conscious, nor do they have wills, free or otherwise, except via the aggregate wills of their human components. They are persons in a purely legal sense, not in the sense that AIs might become. I apply Rabbi Weiss’s model to them via kal vachomer, not because I think the analogy holds.

However, corporations might provide an apt model for considering human responsibility for creating nonperson machines that go of themselves and make morally significant decisions. My sense is that halakhah and Judaism do not yet have adequately developed frameworks for addressing individual responsibility for collective actions with no centralized decision procedure. We need this not just for AIs, but for climate change, labor-capital relations, and much more.

An immediate question specific to AIs is the extent to which we ought to be willing to devolve morally vital decisions to them. I’ll raise one specific kinds of issue here, and look forward to continuing the conversation. My question is whether there is any moral advantage to having humans make direct moral decisions even if an AI would make morally superior decisions. Suppose, for example, that a force of robocops would shoot fewer innocent people than the current force of human cops, with all other law enforcement outcomes remaining stable. Would there be any reason to allow or insist on maintaining a human force?

Here is one possible reason. Asimov accustomed us to think of AIs as rulebound and deductive on moral issues. But machine learning now functions very differently. AIs can be trained to make decisions by imitating  human behavior as recorded in a cache of data, or else by studying the effects of their own past decisions, without necessarily following or developing any abstractly articulable rules. The result of outsourcing practical moral decision making to AIs may be the diminution of practical moral reasoning as a human experience. 

In the context of campaign finance, I used the halakhic principle “it is more of a mitzvah to do the act oneself than to do it via agency” to argue that people should not devolve their political responsibilities and influence to corporations. A somewhat similar argument may apply here. 

Another example: Imagine if there were (there may already be) an app that can decide the colors of stains for niddah purposes more accurately than the vast majority of current rabbis or even yoatzot; the app is trained, let us say, by shimush with Rav Yaakov Neuberger for many years. So on average women will get more accurate results using the app than by asking the rabbis available to them, with much greater convenience, and much less chance of embarrassment. Over time, we might lose all possibility of human psak because the app can’t explain its decisions other than in terms of a prediction of what Rav Neuberger would have decided. Should that be too great a loss, even though the results are better?

I hope these preliminary thoughts contribute positively to this developing conversation. Feedback and specific challenges and questions are of course welcome. I look forward to continuing learning from each other another year of learning together! 

---

[1]  Yakhin note 32 to Mishneh Kilayim 8:5; cf. Edgar Alan Poe’s “The Murders in the Rue Morgue”. See now also the sources collected in Moshe Goldfeder, “Not All Dogs Go To Heaven: Judaism and Beastly Morality”.

Halakhah at Hogwarts

by Rabbi Aryeh Klapper

Dear Friends,

Science/fantasy fiction can be a serious vehicle and a safer space for exploring contentious real-world issues. They are also at high risk of crossing the line into unserious Purim Torah. I hope to have kept this essay on the right side of the line, and ask your indulgence if I failed.

Dear Rabbi Klapper,

My mother often talks about how much she enjoyed learning at CMTL’s Summer Beit Midrash. She wants me to attend after I graduate Hogwarts. I hope I can! Jewish learning is almost nonexistent here. Meanwhile, I have a halakhic question that she said was the kind you enjoy. Can I use my wand as a chanukiah and light it by saying “Lumos? I think that would be cool.

With all best wishes,

Anthony Goldstein 

Dear Anthony:

It’s wonderful to hear from you. I look forward to reading your SBM application as well!

Now to the substance of your question. I’m afraid there’s a prior question that we need to address – I’m sure your mum expected me to raise it, and maybe that’s why she encouraged you to write to me. Is it permitted to use verbal magic for any purpose? Because if illuminating a wand by saying “Lumos” is forbidden, then using your wand as a chanukah light may be a mitzvah haba’ah baaveirah (= mitzvah-committed-via-trangression) and invalid as such.

This is surely a hard question for you to face. It raises the possibility that attendance at Hogwarts is essentially impossible for an observant Jew, and worse, that there is a contradiction between identifying as a wizard and identifying as Orthodox. What would it mean to be a wizard if you can’t use magic?

As a Modern Orthodox halakhist, I have a strong rooting interest in finding a way out of this thicket. I want Torah to be relevant to the Jews who live within the magical community. I believe generally, basing myself on Rabbi Yosef Dov Soloveitchik’s thought, that human beings are encouraged and even mandated to discover and use all the forces that exist in Creation for good.

And yet, I must acknowledge that the Torah adopts a term for magic user – mekhasheifah – to characterize someone liable for capital punishment. I also recognize that various terms for wizard, such as amgushi, have a very negative valence in Chazal. We need to be open to the possibility that our halakhic investigation may yield answers that profoundly challenge us.

Now that I’ve scared you – I hasten to say that I would not have answered your letter had I thought there were no way you could ever use magic. It is forbidden to shout fire in a crowded theater (let alone “Incendio!”) before the ushers are ready to organize the evacuation, even if there really is a fire. And: “Just as a person is commanded to say things that will be heeded, so too a person is commanded to not say things that will not be heeded.”    

There is however all the difference in the world between seeing the use of magic as a halakhic minefield, to be entered only when absolutely necessary; seeing it as an ordinary human capacity with ordinary potential to be channeled for both good and evil; and seeing it as an extraordinary human capacity that requires an extraordinary halakhic response.

For example: I have heard of rabbis who permit only the use of defensive magic. They argue that banning magic entirely would be unilaterally disarming before the enemy, like refusing to fight on Shabbat. If magical skill requires training, and specifically training while young, these rabbis are fine with sending magical Jewish boys to Hogwarts, at least until there’s a hesder-like alternative.

I don’t think the difference between “defensive” and “offensive” magic is sustainable in practice. A successful “Expelliarmus!” can leave even the most powerful opponent at one’s mercy. Consider the halakhic literature about whether Israel’s preemptive First Lebanon War could be classified as defensive and therefore permitted. So this approach would not significantly restrict your studies.

Still, it makes me deeply uncomfortable. I have trouble defining any human capacity as useful purely reactively, as necessary only because “the sitra achra (=other side) can do magic, too”.    

I’ve also heard of rabbis who justify attending Hogwarts solely on the ground that you would otherwise become a danger to yourself and/or others. That seems to me insufficient, terrifying, and profoundly wrong. Insufficient, because what if it turns out halakhically that the use of magic is absolutely forbidden to the point of being yehareg v’al yaavor (= die rather than transgress) as penumbral (= abizraihu) to avodah zarah (= worship that is alien to Judaism)? Terrifying, because it suggests that the world is full of human time bombs, especially in cultures intolerant of magic. Profoundly wrong, because it creates a class of people who are told to view a central aspect of their being as bediavad, and the way they live their lives as justifiable only because the alternative is worse.

Here I need to insert a caveat. Popular treatments suggest that while some magical people are more magical than others, that there is a bright and unpassable line between magical and not. But I have not seen a sufficiently rigorous investigation of this claim.

What if a significant percentage of apparent muggles have latent magicality, perhaps a capacity that develops only after puberty is complete? What if some magical children must be trained lest they become dangerous, but others can live ordinary muggle lives, and there were a reliable way to tell them apart? What if magicality can be acquired, transmitted, or “cured”? The popular belief in immutability is convenient for those who support our status quo in which muggles deny the existence of magics, and magics believe that they cannot live as muggles. That isn’t evidence against its truth. But we should be hesitant to impose radical burdens on either people or on halakhah without exhaustive diligence.       

My own position solves all these difficulties almost magically. I simply don’t believe that what you do is the sort of thing that halakhah forbids as kishuf etc. Rather, I believe that it is a technology that muggle science happens not to understand yet. Forbidding it as magic would be silly, like forbidding the use of electricity would have been several centuries ago. To clarify: I follow Rambam in holding that all effective technologies are inherently permissible, because none of them draw their effectiveness from powers other than G-d.

That leaves the question of what the Torah forbids as magic. I think that according to Rambam, there is a Biblical prohibition against using a technology while believing that it draws its power from a source other than G-d, and a Rabbinic prohibition against pretending that an action of yours drew its power from such a source, especially if the purpose of your pretense is to convince other people that such powers exist.

If you agree with my reading of Rambam, then you can accept your own magicality as a halakhically normal human variation. You may legitimately feel a religious imperative to learn and understand Hogwarts magic as well as you can. But I don’t want to pretend that there are no costs involved.

First, you really have to believe that there is no supraphysical source of power other than G-d. For example, if you’ve previously been convinced that demons are real, you’ll have to unconvince yourself. This may feel like cheating. It is fair and legitimate to worry whether beliefs should be adopted because the alternative is a religious or psychological quagmire, or whether a position affirmed on that basis really counts as a belief.

Second, you know that Rambam’s position has been vociferously rejected by many Orthodox rabbis in good standing, and those rabbis plausibly claim that the Talmud supports them. Indeed, some traditional commentaries refuse to accept that Rambam held his position across the board. So you’ll have to make a claim of certainty about an issue that is highly controversial within our tradition and the Orthodox community. That can feel (and be!) arrogant. Expressing your certainty in public may generate harsh responses.

Third, and almost as hard, you may find yourself thinking much less of people you were otherwise inclined to deeply respect and care about. Their positions on these issues may now engender pity or revulsion instead of being dismissible as quaint and unimportant.

In other words, I am certainly not paskening you a rose garden. At the same time, I don’t want to create the misimpression that all really hard moral challenges to halakhah can be resolved this straightforwardly, or that honest resolutions will always be discovered by the people in your time, place, and community. Sometimes there may be no resolutions available, or the only available resolutions may feel (and may be!) dishonest.

But if you accept my position above, there is no issue of mitzvah haba’ah ba’aveirah in using a wand as a Chanukah light.

Many other issues need to be addressed, such as whether magical speech counts as halakhic action, or whether magical fire is sufficiently parallel to ordinary fire (although on Chanukah perhaps a parallel to miraculous fire is sufficient). Those questions seem to me of public interest, so I’ll post answers on my Facebook page – please have a look! In the meantime, I’ll take that easy way out of suggesting that you make a berakhah over an ordinary oil chanukiah, and use the wand only as a supplemental means of publicizing the miracle.

With all best wishes, and hoping to learn together soon.

Aryeh Klapper

Should What Happens in Beit El, Stay in Beit El?

by Rabbi Aryeh Klapper

Is tranquility a religious goal? Rashi on the opening of Parshat Vayeshev sees desire for tranquility as a recipe for disaster.

Yaakov was yoshev in the land of his father’s being gar, in the Land of Canaan

וַיֵּ֣שֶׁב יַעֲקֹ֔ב בְּאֶ֖רֶץ מְגוּרֵ֣י אָבִ֑יו בְּאֶ֖רֶץ כְּנָֽעַן.

The verbs yashav and gar each refer to some sort of staying-in-a-place. It is conventional to argue that vayeshev refers to a more permanent and committed sort than gar. This argument may undergird or derive from Rashi’s comment:

Another midrash on this:

He was yoshev =

Yaakov sought to be yoshev in tranquility;

there leaped upon him the rogez of Yosef.

The righteous seek to be yoshev in tranquility –

The Holy Blessed One says:

Is it not sufficient for the righteous what is established for them in the Coming World,

that they seek to be yoshev in tranquility in This World?!

ועוד נדרש בו

=  וישב

;ביקש יעקב לישב בשלוה

.קפץ עליו רוגזו של יוסף

– צדיקים מבקשים לישב בשלוה

:אומר הקדוש ברוך הוא

,לא דיין לצדיקים מה שמתוקן להם לעולם הבא

?!אלא שמבקשים לישב בשלוה בעולם הזה

Rashi associates yoshev with tranquility, and by implication gar with a lack of tranquility. Yaakov’s ambition to be yoshev is contrasted invidiously with Yitzchak’s acceptance of being only gar.

The obvious problem is that Yitzchak himself is twice described as yoshev. In Genesis 26:6 we read

וישב יצחק בגרר

Yitzchak was yoshev in Gerar.

even though Hashem has just told him (26:3)

Be gar in this land.

גור בארץ הזאת

Yet Netziv (late 19^th -20^th century) is the only commentary I’ve found that applies Rashi’s criticism to Yitzchak:

Yitzchak was yoshev –

He sought to be yoshev in tranquility rather than in gerut as Hashem had said: be gar!

Therefore he was expelled by/from Avimelekh.

– וישב יצחק

,ביקש לישב בשלוה ולא בגירות כדבר ה’ גור

.מש”ה נגרש מאבימלך

Maasei Hashem (16^th century) explicitly rejected Netziv’s approach and denied that Yitzchak was wrong to be yoshev in Gerar. But his deeper problem is with Rashi: Why was Yaakov wrong to seek tranquility, to the point that he was punished severely by losing Yosef for many years?

Maasei Hashem replies that a close reading of Rashi suggests that Yaakov was not punished at all. “There leaped upon himthe rogez of Yosef”, rather than “there came upon him”, means that Yosef’s fate was already sealed; Yaakov’s desire for tranquility just made that fate materialize earlier. Why?

Because as the Sages said (Shabbat 89b):

Yaakov was deserving of being brought down to Egypt in bronze chains,

but his merit caused him (to instead be brought down to Egypt with full honor).”

G-d moved Yosef’s suffering forward to ensure that Yaakov experienced the tranquility he sought.

I find Maasei Hashem utterly unconvincing, even unreasonable. Nothing in the verse or in Rashi suggests that the tranquility Yaakov seeks is insurance against a hypothetical event many years in the future, with no regard for the decades of suffering in the interim while presuming that his favorite son is dead.       

Surprisingly, neither Netziv nor Maasei Hashem cites an earlier verse in which Yitzchak is yoshev, namely Genesis 25:11:

It was after the death of Avraham

Hashem blessed Yitzchak his son

Yitzchak was yoshev near B’er Lachai Ro’i.

וַיְהִ֗י אַחֲרֵי֙ מ֣וֹת אַבְרָהָ֔ם

וַיְבָ֥רֶךְ אֱ-לֹהִ֖ים אֶת־יִצְחָ֣ק בְּנ֑וֹ

:וַיֵּ֣שֶׁב יִצְחָ֔ק עִם־בְּאֵ֥ר לַחַ֖י רֹאִֽי

This seems to me conclusive evidence that to be yoshev is not intrinsically wrong.

I also wonder whether yoshev actually refers to tranquility. Dr. Joshua Berman argued (in a dvar torah at Yeshivat Har Etzion many years ago) that B’er Lachai Roi represents Yitzchak’s struggle with his father’s legacy of expelling Hagar and Yishmael. Our verse also takes place immediately after a father’s death – Yitzchak dies at the end of Chapter 35, and only the genealogy of Esav (who is yoshev in Har Seir – 35:8) separates that death from vayeshev Yaakov. Maybe Yaakov is yoshev where his father was gar because he too was working through legacy issues.

Nonetheless, I am fascinated by the way Maasei Hashem counterreads Rashi. In his version, there is nothing wrong with seeking tranquility, and in fact G-d rearranges His plan of history in order to accommodate Yaakov’s desire for tranquility. This counterreading requires us to repunctuate Rashi. G-d is no longer expressing ironic outrage. Rather, he is acknowledging the legitimacy of Yaakov’s desire for tranquility in This World over and above the reward awaiting him in the Coming World. (Note that the Chassidic commentary Bat Ayin adopts the same approach as Maasei Hashem.)

The early 20^th century Responsa Lev David (#88) shared Maasei Hashem’s discomfort with Rashi’s apparent moral, and found an astonishing way to avoid it.

I always had difficulty with what Rashi said in Vayeshev . . . until I found in a book that this was a printer’s error. Originally it was written הש’, meaning “The Satan”, but the printer ‘corrected’ it and wrote that השם/Hashem said, and afterward they wrote השי”ת/Hashem the Blessed, and then they printed (as we have in Rashi) “Said the Holy Blessed One”. And I found a support for his words in the Midrash to Parshat VaYeshev: “The Satan comes and makes charges and says: “Is it not enough for the righteous etc.”

In fact, all the midrashim I’ve found have the Satan rather than G-d making the critique of Yaakov. I wonder whether Rashi deliberately emended, and if so why.

Leaving Rashi aside, I wonder whether Lev David is correct that attributing the critique to the Satan means that we need no longer take responsibility for it. Here as in the Book of Job, the Satan cannot act to harm without G-d’s permission.

In Halakhic Man f.n. 4, Rabbi Yosef Dov Soloveitchik endorses Rashi as-is, with gusto. Yaakov’s desire for tranquility reflects a culpable unwillingness to do the work necessary for true religious growth. It indicates a worldview in which it is always time for shaloshudus somewhere, so that we can sing the religious fantasy of Psalm 23 continually to avoid facing the challenges of the social and intellectual world. 

The Rav’s approach was utterly vital to me as a teenager, and I still find it highly attractive. But I wonder if it is a universal truth. Maybe for some people tranquility is culpable avoidance, and for other people it is the best way to fully develop their spiritual and interpersonal capacities.

I’ve read Genesis that way for a long time. Avraham is the man of nisayon, who grows via dramatic tests; Yitzchak is a man of stability; Yaakov wants to be his father, but in fact needs to be his grandfather. Yaakov tries to forestall nisyonot by making his obligations to G-d conditional on tranquility; G-d thwarts him by fulfilling the conditions without giving him tranquility.

Midrash Tanchuma launches a fierce critique of Yaakov that I think works on the same basis:

Come and see:

When Yaakov went to Aram Naharayim, what is written there? Yaakov vowed a vow etc.

G-d responded to every clause.

He became rich and was yoshev, but did not fulfill his vow –

so G-d brought Esav upon him who sought to kill him, and he took from him all that bribe . . .

Yaakov did not realize – so G-d brought the angel upon him to wrestle with him but not kill him . . .

Yaakov did not realize, so the tragedy of Dinah came upon him . . .

Yaakov did not realize, so the tragedy of Rachel came upon him . . . 

Said The Holy Blessed One:

How long will this righteous one suffer and not realize which sin he is suffering for?! I will tell him, as Scripture says: Hashem said to Yaakov: Arise, go up to Beit El and be yoshev there. (35:1)

G-d never tells Yaakov to leave Beit El. Yet in 35:16 they travel from Beit El. Perhaps Yaakov’s sin is that he was yoshev in all the wrong places, that he was constantly seeking to be his father rather than remaining where his own religious faculties reached their apex.

Shabbat shalom!

The Duel in Penuel

By Rabbi Aryeh Klapper

“These words contradict Scripture (and) are forbidden to hear, all the more so to believe”. It was a transgressive thrill to read that passage of Ramban (Bereishis 18:1) in tenth grade. Did he really just say that about Rambam? Yes, he did. Rishonim really talked that way about each other’s ideas. If Ramban could think Rambam had strayed so far from truth, surely if was ok for me to think the same of my teachers on occasion, or even to side with Rambam against Ramban here.

Ramban was reacting to Rambam’s statement in Guide to the Perplexed 2:41 that any Torah narrative in which a human being sees or hears an angel must be either a dream or else a prophetic vision. (Presumably the same is true if someone feels or smells or tastes an angel.) Guide to the Perplexed 2:42 explains that this is so even if the human being does not recognize the angel as such, and even if the Torah never explicitly identifies the angel as such. So long as it actually was an angel, the story must be a vision or dream.

Rambam argues that readers sensitive to literary structure should recognize that the Torah often tips its hand that what follows will not take place in the physical world. For example, Parashat Vayeira begins “And Hashem appeared to him . . . he saw, and behold three men”. This is an example of the literary device klal uprat, in which a broad introductory statement is followed by a detailed exposition. Avraham’s seeing of “three men” and all that follows (at least) until Avraham gets up early in the morningafter Sodom’s destruction, is the exposition of how Hashem appeared to him. Similarly, Bereishis 32:2

ויעקב הלך לדרכו ויפגעו בו מלאכי א?להים

and Yaakov went on his way, and angels of G?d encountered him

is a broad introductory statement that subsumes Yaakov sending angels to his brother Esav, and the ominous report they returned. Rambam contends that it covers everything up to and including the end of the wrestling match.

   
How do we know that Yaakov’s wrestling partner was an angel? Rambam states that “it becomes obvious at the end”. I’m not certain how. Because Yaakov demands a blessing? Human beings give each other blessings throughout Tanakh, and of course Yaakov and Esav’s relationship fractured over such a blessing. Because the loser refuses to give his name? Many rishonim contend that it is impolite to ask a vanquished foe for their name, for the same reason that this happens all the time in Arthurian stories – it allows the victor to re-humiliate the loser every time the story is retold. Most likely Rambam is referring to the content of the blessing – because you have striven with Elo?im and with men – and Yaakov’s naming of the location – because I have seen Elo?im face to face, and my nefesh was preserved. But the word El?him can just mean “mighty”.

Ramban counters that all these narratives seem to have direct consequences in the physical world. Sodom is really destroyed, and Yaakov limps away from the wrestling match. How can purely mental events effect such consequences?  

(If Rambam is correct), I don’t know why (Yaakov) was limping on his thigh when awake . . .

According to this opinion of his, we’d have to say the same about the matter of Lot, that the angels never came to his house, and that he never baked matzot for them that they ate, rather it was all a vision. But even if he raised Lot to the level of prophecy, how could the evil sinning people of Sodom be prophets? yet otherwise, how would they know that men had come to Lot’s house? If (you say that the entire story” is Lot’s prophetic vision . . . then Lot should have remained in Sodom (during its destruction, and be dead)?! So (Rambam) must think that all these things happened on their own (i.e. naturally), but the statements (in the Torah about them) are all visions.

Rambam’s partisans respond that Yaakov limped because mental events can have physical effects on one’s own body, and that Avraham’s vision was a religiously filtered experience of the actual destruction of Sodom; he perceived e.g. a volcanic eruption as G-d making it rain sulfur and salt. I don’t know why Ramban thinks that this understanding of Avraham’s vision “contradicts Scripture”. (Sara Krishtul correctly wonders why, in Ramban’s critique, Lot could not have escaped Sodom in a time between his vision and the actual destruction.)

Furthermore, I suggest tentatively that Ramban misunderstands Rambam in a crucial way. Rambam does not think that the people of Sodom, including Lot, have any visions. They are all just characters in Avraham’s vision. Similarly, the “real” Esav never encounters Yaakov’s messengers.

I also don’t understand why Ramban sets the stakes of this disagreement so high. Immediately after his anti-Rambam flourish, he too insists that angels are not part of the physical world:

In truth,

everywhere that Scripture mentions the seeing of an angel, or the speech of an angel,

it is in a vision or dream,

because the senses cannot capture angels.

ובאמת

– כי כל מקום שהוזכר בכתוב ראיית מלאך או דבור מלאך

,הוא במראה או בחלום

.כי ההרגשים לא ישיגו המלאכים

What differentiates this position from Rambam’s? Ramban spends considerable ink explaining the difference between prophetic and non-prophetic visions and differentiating among types of angels. But I still do not understand why people are more likely to emerge limping from nonprophetic rather than from prophetic visions, or for that matter the difference between angels that appear in visions and those that are

,כבוד נברא במלאכים

,”יקרא אצל היודעים “מלבוש

.יושג לעיני בשר בזכי הנפשות כחסידים ובני הנביאים

Glory created as angels

called by the cognoscenti “Attire”

perceptible to flesh-and-blood eyes of the pure of soul like the pious and the students of prophets

Ramban’s frustrating last words on the subject are

ולא אוכל לפרש

I am unable to explain.

This is not I think a statement of humility, but rather a claim about the (lacking) spiritual stature of his readership – one completely accurate with regard to this reader. Perhaps this is poetic justice for my youthful celebration of his sharpness against Rambam, which, to be fully honest, is not wholly in the past.

I have long been partial to Nechama Leibowitz’s understanding, and Howard Fast’s, that angels who wrestle are aspects of the self that are not fully integrated. This is the most powerful understanding of the midrashic reading that the angel in our story is the avatar of Esav. Aviva Zornberg brilliantly contends that Yaakov has been wrestling with a sense that his impersonation of Esav was as much self-discovery as impersonation. This seems to require an angel with a purely psychic manifestation.

The contrasting position was Rashbam, who seemed to require a physical angel.. Rashbam argues that the angel came in order to prevent Yaakov from running away, and therefore left at dawn when Yaakov could no longer flee successfully. I’ve said some harsh things about Rashbam’s position. I didn’t declare it forbidden. But his pure instrumentalization of the angel seemed to me close to contradicting Scripture, which imho clearly gives the angel a persona.

However, Rabbi Aviad Tabory gave a shiur at YI Sharon several weeks ago that opened up a possibility I had overlooked. Perhaps Yaakov intended to flee not because he was afraid of losing the battle, but rather because he was afraid that winning would cement his identification with Esav. He wrestles with this challenge – procrastinates? – long enough that fleeing is no longer an option. This understanding makes Rashbam compatible with the position that the angel was internal to Yaakov.

My question then is whether the issue of angelic materiality – really how many angels can dance on the head of a pin – has any implications at all for the meaning of Chumash, and if it has none, why Ramban chose to foment this tempest in this teacup. Suggestions and corrections are welcome.

Shabbat shalom!

Dinah and DNA

by Aryeh Klapper

Midrash Halakhah is no longer a productive discipline. We no longer use either Rabbi Yishmael’s (כלל ופרט) or Rabbi Akiva’s (ריבוי מיעוט) formal exegetical tools to derive law from Torah. But that doesn’t mean that we no longer derive halakhah from Torah. It means only that we do so only in less formulaic ways.

Our understanding of past midrash halakhah is often predicated on a sharp conceptual distinction between “midrash” and “asmakhta”. “Midrash” produces law with “deoraita” status, while “asmakhta” produces law with “derabanan” status. While each term in those statements should be extensively qualified, the framework generates a useful corollary, namely: When a “midrash” is rejected as an incorrect interpretation of Scripture, any legal assertions associated with that midrash are rejected as well unless/until they can be derived from another Scriptural source. Rejecting an “asmakhta”, by contrast, has no formal impact on the acceptance of associated legal assertions.

This has a practical consequence that may seem ironic. A law that claims to be deoraita is often more vulnerable to argument than one that claims to be derabanan.  

Is the distinction between “midrash” and “asmakhta” meaningful in a post-Midrash Halakhah world? I think that question deserves extensive treatment.  For now, here is one parallel irony to consider: A source based on a feeble interpretation of Scripture may be a stronger halakhic precedent than one based on a strong argument. This is because the feeble connection to Scripture suggests that the source holds the underlying position strongly enough to use it as a frame for understanding Scripture, whereas the strong connection may reflect only a literary preference.

It’s important to recognize that even the most compelling Scriptural interpretations usually generate only abstract halakhic positions. Applying those positions to our reality is an independent step, and it’s entirely possible for such positions to have no application in our reality. This is especially true for positions that are derived indirectly from interpretations that are valuable for their non-legal implications, for example because they help us better understand a narrative. 

The Torah’s account of the coming-into-being of Dinah generates a host of interpretations with no direct legal intent. Nonetheless, such interpretations are cited as precedent in at least five ongoing halakhic conversations. In each of these contexts, it might be important to clarify the extent to which the legal force of the precedent depends on how compelling we find the interpretation, or conversely, that the legal force of the precedent is enhanced by the weakness of the textual connection.

The five halakhic contexts are:

1. Whether/when it is permissible to pray for a child to be of a particular gender
2. Whether/when to test fetuses for inevitably fatal conditions 
3. Whether to cite shehechiyyanu for the birth of a daughter
4. Which of the egg mother, the gestative mother, and/or the parturitional mother is the halakhic mother in a case of surrogacy
5. Whether the Torah recognizes the possibility of a human being changing gender

I’ll briefly present the interpretational frameworks that generate the legal precedents; explain how they generate the legal precedents; and then leave you to think about what force those precedents have and why, and whether/how they should and should not be applied to our reality as we perceive and understand it.

The Torah uses a four-verb formula to introduce ten of Yaakov’s twelve sons:

1. became pregnant = הרה;

2. she bore (a child) = ילד;

3. she said =אמר ;

4. she called – קרא.

“She said” introduces an etiology for the child’s name, while “she called” introduces the name itself.

For example:

Bilhah became pregnant/ותהר

She bore/ותלד to Yaakov a son

Rachel said/ותאמר: Elokim has judged me, and also heeded my voice, and He gave me a son;

Therefore she called/קראה his name Don.

Only two of those verbs are used to introduce Dinah, Yaakov’s only named daughter.

And afterward, she bore/ילדה a daughter

and she called/ותקרא her name Dinah.

Many pshat commentators feel compelled to explain why the Torah doesn’t include Leah’s speech at Dinah’s naming. For example, Rashbam apparently asserts that there was no speech – people acknowledge G-d less for the birth of daughters than sons. I don’t understand how this (and Ibn Caspi’s balder statement that women are worth less than men) solves the problem. Dinah was given this name; even if it wasn’t intended as acknowledgement to G-d, why is understanding “Dinah” less important than understanding “Don”?

Another suggestion is that Dinah was named retroactively, after she was raped, rather than on the basis of her parents hopes and dreams. I’m not clear on how that explains her name. Maybe Dinah is just piggybacked on the explanation of Don, since they come from the same root.

Targum Yonatan fills the structural gap with what seems to be an entirely invented statement:

ומן בתר כדין ילידת ברת וקרת ית שמה דינה,

ארום אמרת:

“דין הוא מן קדם י”י דיהון מיני פלגות שיבטייא, ברם מן רחל אחתי יפקון תרין שיבטין היכמה דנפקו מן חדא מן אמהתה.”

ושמיע קדם י”י צלותא דלאה, ואתחלפו עוברייא במעהון,

והוה יהיב יוסף במעהא דרחל, ודינה במעהא דלאה.

Afterward, she bore a daughter, and she called her name Dinah,

because she said:

“It is a din from Hashem that half the tribes should come from me, and that from Rachel my sister two tribes should emerge, just as (two) emerged from each of the maidservants.”

Leah’s prayer was heard by Hashem, and the fetuses in their wombs were switched,

so that Yosef was in Rachel’s womb and Dinah in Leah’s womb.

This insertion may be derived from the other structural gap we identified, the absence of “she became pregnant”. The author of this interpretation concluded that Leah did not ‘become pregnant’ with a daughter, but rather with a son, although she gave birth to a daughter.

An obvious problem is that the Torah says regarding Yosef She became pregnant, and she bore a son: if Yosef was switched with Dinah, then Rachel did not become pregnant with him?! This seems an unanswerable weakness. However, there is a way to save the approach, if not the specific interpretation. The Torah does not say that Rachel ‘became pregnant’ with Binyamin. So perhaps Dinah was actually switched for Binyamin. This requires significantly revising the timeline of the verses, so that Dinah is born after Yosef, and after Yaakov leaves Lavan’s house, despite being introduced before then. However, this makes sense structurally in that we finish listing the children of Leah, Bilhah, and Zilpah before beginning Rachel’s, and provides an excellent explanation of the Torah saying afterward she bore a daughter. It may also require us to understand the narrative of Yaakov’s decision to leave Lavan as fundamentally about Rachel. This may be a ground for the position that Rachel dies because Yaakov could not be married to two sisters in the Land.

We can account for the absence of she became pregnant regarding Dinah without needing a parallel son by saying that she was not switched in the womb for a male fetus but rather transformed in the womb from a male fetus. This is the position taken by Rav on Berakhot 60a.

מיד נהפכה לבת

Immediately she was transformed into a daughter.

Note also that the calculation attributed to Leah is likely derived from yet another structural oddity. Yosef’s name is apparently given two separate etiologies:

וַתַּ֖הַר

וַתֵּ֣לֶד בֵּ֑ן

וַתֹּ֕אמֶר: אָסַ֥ף אֱ-לֹהִ֖ים אֶת־חֶרְפָּתִֽי:

וַתִּקְרָ֧א אֶת־שְׁמ֛וֹ יוֹסֵ֖ף לֵאמֹ֑ר: יֹסֵ֧ף יְקֹוָ֛ק לִ֖י בֵּ֥ן אַחֵֽר:

She became pregnant

She bore a son

She said: Elokim has gathered away my shame

She called his name Yosef, saying: Hashem will add to me another son.

The second etiology uses the verb-root of Yosef/will add, but seems to make no sense: why would Rachel react to Yosef’s birth by asking for another son?  This can be resolved by saying that Rachel knew that she would produce two tribes.

The last word of the second etiology, אחר/another, is consonantally identical with the אחר/afterward that introduces Dinah’s birth. This supports a connection to Yosef rather than Binyamin.

Note also that so far as I can tell all commentators on the Talmud assume that even if the fetuses were not switched in the womb, Rachel and Leah were simultaneously pregnant, and therefore if Dinah was gender-switched from male to female in the womb, Rachel’s son must similarly have been switched from female to male. Rav Yaakov Kaminetzky suggests that the transformation was not absolute, so that Dinah retained masculine traits, and Yosef female traits.   

However, this entire midrashic edifice still hangs from a hair, since there is a non-miraculous way of giving significance to the absence of she became pregnant, namely that Dinah was a twin and conceived as part of the same pregnancy as Zevulun. Moreover, it’s not at all clear to me that the absence of she became pregnant is significant. I began this essay by noting that the Torah uses all four verbs to introduce 10 of Yaakov’s sons. The two missing are Zilpah’s children, and the Torah mentions neither pregnancy. So perhaps there is no significance at all to whether or not it is mentioned, or perhaps we’ve completely mistranslated the verb and it means something else entirely.

So much for the interpretational framework. Some halakhic connections are:

1. Leah successfully prayed for the gender transformation of her fetus. Therefore, it is not considered a “vain prayer”, meaning a prayer that seeks to alter the past, if one prays for an already conceived child to be a particular gender. Note that the Bavli and Yerushalmi dispute how late in the pregnancy this might be possible, and that Rav Yaakov Kaminetsky’s interpretation fits interestingly with contemporary scientific accounts of how a person can be genetically of one sex and yet apparently develop physically as the other.
2. Chashukei Chemed, which I want to emphasize again is not a guide for practical halakhah, suggests that it is best to pray for a fetus’ health before any tests are done, as prayer is less likely to be efficacious in changing an already observed past (shades of Heisenberg?), and Leah’s prayer demonstrates that fetuses are wholly plastic to prayer.
3. Rabbi Moshe Steinberg of Kiryat Yam used the absence of an etiology for Dinah’s name as evidence against Tzitz Eliezer’s position that one makes Shehechiyyanu for the birth of a daughter. Tzitz Eliezer’s impassioned response is worthwhile reading (Tzitz Eliezer 10:21:2).  
4. On the reading that Dinah was switched with Rachel’s female fetus, since the Torah seems to clearly identify Dinah as Leah’s daughter, it follows that genetic motherhood is halakhically irrelevant. (Of course, one can argue that nothing can derived from pre-Sinai narratives, that miracles cannot be used as precedents, etc.)
5. On the reading that the fetuses were transformed rather than switched, then plainly halakhic sex is not immutable at the moment of conception, or alternatively, the gender of a soul does not irrevocably determine the sex of a body. (Of course, this has not necessary implications for post-birth transformations, artificial transformations, and so forth, nor does it set forth any criteria for what constitutes such a transformation.)

Shabbat shalom!