When looking at Rivash, it is quite apparent that there are series of responsa, sometimes quite lengthy, that appear to be part of a single correspondence. Take, for example, responsum #254: It is addressed to Rabbi Pinhas ben Shalamya in Xativa, and the next eight responsa all begin with the words “עוד שאלת”, “you also asked”. In fact, well over half of the teshuvot in the collection begin with such formulations.
Sometimes the series are numbered. Responsum #400 was addressed to Rabbi Moshe Maskaran in Huesca. It begins: “The first question…” The next three responsa, begin, respectively, “The second question”, “The third question”, and “The fourth question”. The first three responsa in this series of four questions relate to the laws of Pesah; was he cracking wise about the “four questions”? Perhaps. Either way, it is apparent that these four responsa were part of a single correspondence. It stands to reason that formulations like “you further asked” indeed preserve the original order of the responsa, at least within each series. Our inclination was therefore to conclude that all nine of the responsa from 254 to 262 were to Xativa, and plot the map accordingly. Likewise for all series.
On the other hand, we saw with respect to R. Moshe Feinstein that one correspondence could be chopped in two and published in different places, even different volumes. The responsa of the Rishonim, moreover, are sometimes chopped up and scrambled into all different orders. These texts circulated in manuscripts whose orders were not uniform. In his work on on the various responsa collections of Rashba and Maharam Mi-Rothenberg, Prof. Simcha Emanuel takes on the massive project of cross-references the numbering schemes of the various collections and citations. Moreover, we noticed relatively early that Rivash’s responsa to North Africa appear at the beginning of the collection of responsa, even though he lived there only at the end of his life. The responsa seemed to be in reverse chronological order. This indicated that the order had been tampered with, and that we could not therefore trust that what seemed like series of responsa to one individual were in fact so.
There are no accessible digitized versions of any complete Rivash manuscript (at least not on KTIV – yet), so that was no help.
The first bits of resolution came from the introduction to the Machon Yerushalayim edition of Teshuvot Ha-Rivash, which records that R. Yosef Karo was working from a manuscript at one point, and only later obtained the print version of Teshuvot Ha-Rivash. Moreover, his manuscript was incomplete and contained only about one-third of the responsa in the collection.
We also discussed the question with Prof. Emanuel, who graciously provided us with several sources that discuss the issue. The most extensive and important one appears in Abraham Hershman’s work on Rivash. Hershman provides a full account of the manuscript and print history of Rivash’s responsa. He cites evidence of some mistakes in attribution (which we fixed in the latest update to the map), but confirms that, all in all, series of responsa are reliably preserved intact. Here’s the English version of Hershman’s book. The relevant discussion begins on p. 5:
Hershman gives an important accounting of the transition from manuscript to print. Rivash “published” his responsa in three different volumes, which circulated separately in manuscript prior to the first printed edition. These three volumes were from different periods of his life: One from the beginning of his career, mainly in Barcelona and Zaragoza (Vol I); one from the second part of his career, when he was in Zaragoza and later Valencia (Vol II); and one from the third part of his career, in North Africa (Vol III).
These three volumes were printed in reverse chronological order. Volume III corresponds to responsa 1-186; Volume II to responsa 187-359; and Vol I to responsa 360-518. Hershman and the Machon Yerushalayim intro both show places where a manuscript responsum of the Rivash is cited, but its number does not match the numbering in the printed editions – but does correspond to the numbering within the particular volume. For instance, Responsa Mabit #32 refers to Responsa Rivash #12. It’s not #12 in the printed editions – it’s #198, which is Vol II, #12 (198-186=12). This threefold partition seems to have originated with Rivash himself. He seems to have ordered each one, albeit not necessarily chronologically, but he keeps series intact. In fact, he did not necessarily think of them as separate responsa; it was later readers and copyists who presumably broke up these lengthy responsa for easier reference. Moreover, the three volumes, if they are not ordered internally by chronology, are from different periods of his career.
This gives us a pretty good picture of the multi-stage process by which Rivash’s correspondence was transformed from individual letters, into collections (by the author himself), to authoritative sources (by Beit Yosef, Mabit, and others who cite him, as reflected in the numbering of manuscript responsa), to a single printed volume.
Though we have a few quibbles with Hershman (תנס is Tenes, Algeria, not Tunis, Tunisia, for example), in all, his work is an unparalleled resource on Rivash, and through it we were able to refine and confirm many of the data points on the Rivash map. It has reinforced for us that much have our path has been paved by scholars who accomplished incredible research with a fraction of the tools and none of the computing power we have today.
We have updated the Rivash map accordingly, including a color-coding of the original three volumes. It bears out what we have seen in the past. Vol I, representing the early part of Rivash’s career, covers mainly just Aragon and Navarre. In Vol II, written during the peak of his career in Iberia, there are many more responsa to Castille in addition to Aragon and Navarre. Almost all of Vol III is to communities in North Africa, and the few exceptions are mainly to relatives who were still in Aragon. We suggest opening it in its own tab.
It is worth asking why he did not write any more than a handful of responsa to Iberian communities after his move to Algiers. It’s worth noting: volumes I and II are about as geographically similar to each other as two poskim from the same tradition tend to be. This isn’t necessarily so surprising, since the geographical spread of a posek isn’t much more predictive than the spread of his area (for example: an earlier volume of Hatam Sofer will be only somewhat more similar to a later volume than Ktav Sofer might be). However, volume III is as different from I and II as anything we’ve seen — it’s truly apples to oranges. We’d see similar numbers comparing German poskim to Galitzianers, and it suggests that to a substantial degree, Rivash “changed jobs” with his move to North Africa.