Friday, October 17, 2008

ish and isha

The first words the Torah quotes Adam as saying appear right after his wife was created. He gives her a name, but also mentions a connection between his name and hers:

וַיֹּאמֶר, הָאָדָם, זֹאת הַפַּעַם עֶצֶם מֵעֲצָמַי, וּבָשָׂר מִבְּשָׂרִי; לְזֹאת יִקָּרֵא אִשָּׁה, כִּי מֵאִישׁ לֻקְחָה-זֹּאת.

Then the man said, "This one at last is bone of my bones and flesh of my flesh. This one shall be called Woman (isha) for from man (ish) she was taken". (Bereshit 2:23)

As you might imagine, I love the idea that the first recorded human sentence include an etymology! Ibn Ezra says that the words ish איש and isha אשה are related, and Rashi goes even further. Based on the midrash in Bereshit Rabba (18:4), he says that from this verse we learn that the world was created in Hebrew!

However, there are a few difficulties with what we've said so far. First of all, almost all modern linguists say that ish and isha aren't related. Ish comes from the root אוש, meaning strength (the related root אשש means "to strengthen"), and isha derives from אנש, meaning weak. (The common plural of both - anashim אנשים - "men" and nashim - נשים - "women" also derive from אנש). Aside from the fact that the Torah mentions them together, it might seem difficult to believe that they aren't from the same root. We see the letter heh added as a feminine suffix in many words. Why shouldn't we accept what appears to be the obvious etymology here as well?

Well, first of all, we should be careful of "obvious" etymologies. I've warned a number of times of the danger of assuming that words in English and Hebrew that have similar sounds and meanings are related. But we need to be just as careful when it comes to words in the same language. My friend Mike Gerver has a list of false cognates in English, with pairs of words that most people would find it hard to believe aren't from the same root: for example, pull and pulley, isle and island. I would add to this list one very relevant to our topic: male and female!

Secondly, it's important to note the the woman is receiving a name here. We see many times that the etymology of names given in the Torah does not match up exactly with the linguistic etymologies. For example, the Torah connects the name Noach נח with the root נחם, even though the two aren't connected. (For a good analysis of this phenomenon, read this post on Parshablog.)

But perhaps most significantly, there's a grammatical problem here as well. The word isha has a dagesh in the shin. Ibn Ezra says this is to distinguish it from the homonym ishah -"her husband." Modern linguists, however, have shown that this dagesh is due to the "dropped nun" phenomenon (see other examples here.) In addition, the shin in ish and the shin in isha aren't really cognate. Horowitz writes (page 107):

Strange and unbelievable as it seems the word אשה has absolutely nothing whatsoever to do with the word איש. In אשה in the first place a nun has fallen out; the word is really אנשה (insha). The plural נשים gives some hint of that. The really important fact, though, is that the shin of אשה is really a tav. In Aramaic the word for woman is either אתא or more commonly אתתא.
Klein points out that there is an Aramaic form אנתתא, as well as the Arabic אנתת, which show both the dropped nun and the tav instead of shin.

A number of modern Jewish commentaries deal with the apparent conflict between the verse in Bereshit and the currently accepted etymologies. Sarna writes in the JPS Genesis:

Hebrew 'ishah ... 'ish, though actually derived from distinct and unrelated stems, are here associated through folk etymology by virtue of assonance.
Kil in the Daat Mikra in a footnote (with no apologetics) writes that ish is from אוש and isha is from אשש, whereas the plural of both comes from אנש. Cassuto makes a similar comment, but has isha coming from the root אנת / אנש as we've mentioned above.

Probably the earliest "modern" commentator that I could find who dealt with the issue was Shadal. He wrote:

Moses recorded these words as they were pronounced in his time, even though ishah was not actually derived from ish, but rather from enash, which became enesh (as gevar became gever), yielding the plural anashim, as well as the feminine form inshah, which became ishah. The word ish, however, has been preserved in its original form.
Now while I'm generally willing to explore any resource in order to find the history of words on this site, I was a little nervous going forward here. I needed to present an "unbelievable" etymology, and it seems to contradict a verse in the Torah! And from conversations with friends, I had the feeling that my "modern" commentators weren't going to cut it.

So I was relieved when I found that the Radak in his Sefer HaShorashim (entry אנש) wrote that:

ואפשר שתהיה אשה מזה השרש והדגש לחסרון הנו"ן והראוי אנשה

"And it is possible that isha is from this root (אנש) and the dagesh is due to the missing nun, and it is properly אנשה"
So we have some early grammatical proof. (By the way, I'm not sure if the Radak was the first one to come up with the אנשה theory - it was just the earliest I could find. If anyone has any other sources, I'd be glad to see them.) However, it doesn't entirely satisfy the issue. Rashi's quote from the midrash above still stands - aren't we supposed to learn something about the role of Hebrew from this verse?

Let's look a little closer at what Rashi wrote:

לשון נופל על לשון. מכאן שנברא העולם בלשון הקודש

Most of the English translations of Rashi imply that he's referring to an etymological connection between ish and isha. For example, this site translates it as:

[The words איש and אשה ] have the same root. From this [we derive] that the world was created with the Holy Tongue.
Another site has a similar translation:

One expression coincides with the other [i. e., the words אִישׁ and words אִשָּׁה have the same root]. From here is derived that the world was created with the Holy Tongue.
However, these are not precise translations of Rashi's language. He does not say that the words are from the root - he says "lashon nofel al lashon" - which means a play on words or a pun (literally "one expression falls on another expression"). We find this term in a related midrash in Bereshit Rabba (31:8). The midrash notes that God told Moshe to make a seraph (Bamidbar 21:8). Moshe, however, makes a copper snake - a נחש נחושת - nachash nechoshet (21:9). His reason for doing so was that this was lashon nofel al lashon - the words nachash and nechoshet resemble each other. And here too, the midrash learns from this that the world was created in Hebrew.

I think anyone reading that midrash alone would not assume that nachash and nechoshet are related etymologically (they're not, by the way, but that's for a different post). So perhaps here too, we don't need to assume that ish and isha are related etymologically by virtue of them being lashon nofel al lashon.

I think this point was well explained by Rabbi Yitzchak Arama, author of Akedat Yitzhak (chapter 8). He writes there that ish and isha do not have the same root. He connects ish to strength, but curiously does not present the etymology of isha. He distinguishes ish and isha from other animals where the female is marked by a heh suffix, such as par פר -"bull" and parah פרה - "cow", keves כבש -"lamb" and kivsa כבשה - "ewe", and more. (This is contrast to the Chizkuni on 2:23 who writes that humans are the only species that the terms for male and female come from the same root. How he missed the numerous examples where it does occur, I don't know.)

He then points out something very important about the concept of lashon nofel al lashon. He quotes the midrash that Rashi quoted above, and writes:

ומההכרח להיות כוונתם ז"ל מה שאמרנו שאל"כ מאי קאמרי שהוא לשון נופל על לשון והלא לשון אחד ממש הוא כמו פר ופרה כבש וכבשה

"And this must have been their intention (when they wrote the midrash), as we said (that the words come from different roots). For if not, then what does it mean, 'lashon nofel al lashon' - for is it only one lashon (expression), like par and parah, keves and kivsa?"
So from the Akedat Yitzhak's reading (which I think makes a lot of sense), the midrash and Rashi actually support the idea that ish and isha aren't etymologically related.

One other thing bothered me about the midrash. Not only does it say that the world was created in Hebrew, but it uses the proof from the verse that the Torah was given in Hebrew:

From here you learn that the Torah was given in Hebrew. R. Pinchas and R. Chilkiya in the name of R. Simon: Just as the Torah was given in Hebrew, so too was the world created in Hebrew. Have you ever heard gini ginia, anthrope anthropia, gavra gavreta? Rather ish and isha. Why so? For they are similar sounding words (lashon nofel al lashon).
The midrash is showing how in Greek and Aramaic the words for male and female are not related at all - there are no such words in Greek as gini and anthropia, nor gavreta in Aramaic. (This "proof" doesn't work as well in English, were we have the related "man" and "woman", but it does allow the English translation of Bereshit 2:23 to duplicate the assonance.)

But why would anyone even assume that the Torah was written in another language? I think we need to pay attention to the linguistic pluralism facing the Jews in the land of Israel at the time of the midrash (R' Simon lived in 3rd century Lod.) Both Greek and Aramaic were seriously challenging Hebrew not just as the vernacular, but for religious significance as well. The Aramaic and Greek translations of the Torah were very popular, and we find numerous permissions in Jewish law to pray in other languages. An example of just how far this trend had progressed can be found in the following midrash (Sifrei Devarim 343):

"ויאמר, ה' מסיני בא" (דברים לג, ב) - כשנגלה הקדוש ברוך הוא ליתן תורה לישראל, לא בלשון אחד נגלה, אלא בארבעה לשונות:
"ויאמר, ה' מסיני בא" - זה לשון עברי.
"וזרח משעיר למו" - זה לשון רומי [=לטינית].
"הופיע מהר פארן" - זה לשון ערבי.
"ואתה מרבבות קדש" - זה לשון ארמי.

"The Lord came from Sinai; He shone upon them from Seir; He appeared from Mount Paran; and came from Rivevot Kodesh" (Devarim 33:2)
When the Holy One gave the Torah to Israel, he didn't reveal himself in one language, but in four:
"The Lord came from Sinai" - this is Hebrew.
"He shone upon them from Seir" - this is Roman (Seir is Edom, identified with Rome. This is likely Latin, although some say that it refers to Greek.)
"He appeared from Mount Paran" - this is Arabic (Yishmael lived in Paran, from whom the Arabs are descended)
"And came from Rivevot Kodesh" - this is Aramaic (the word for came, ata, is Aramaic)
It was against this background that R' Simon needed to point out that the Torah was given in Hebrew. I'm sure that most Jews of the time knew that Moshe didn't speak Aramaic or Greek. But the message here is that word play can really only be appreciated in the original language. Something is always lost in translation. Puns, despite them being the "lowest form of humor", are the inside jokes of language, allowing a real connection between writer and reader.

I grew up in a family where we constantly made puns - some of them real groaners. And I still appreciate them today - you've seen me quote the comic strip "Pearls Before Swine". So as hard as it is for me to give up the first quoted sentence by a human as an etymology, I'm even more thrilled to see it was a form of a pun...

Thursday, October 02, 2008

google now translates to and from hebrew!

Thanks to our faithful commenter Joel Nothman, I now know that Google's translation services can now translate to and from Hebrew! It's not perfect (and Joel's blog post points out some of the flaws), but it's an important step forward. I wonder when they'll translate from Biblical and Talmudic Hebrew to Modern Hebrew...