This is the second of a sequence of posts (the first post can be found here), from a draft of a forthcoming monograph by Dr. William Gewirtz that addresses the period of bein hashemashot. Each post briefly summarizes about 20 pages of the monograph and contains 4 - 6 critical pages from that section as well.
The monograph addresses the period of bein hashemashot, the most fundamental area of dispute in the area of zemanim. What is proposed is an astronomically accurate hybrid position between the diametrically opposed conceptual views of the geonim and Rabbeinu Tam. That position justifies, to varying degrees, the practice of countless generations of European Jewry that started Shabbat well after sunset on Friday evening. Though often ignored in modern times, practical equivalents of this hybrid position have had major adherents throughout the generations. Furthermore, our goal is to demonstrate that such a position is not just plausible, but in fact the preferred reading of the primary text of the gemara in Shabbat concerning bein hashemashot.
This second section
- Introduces the major sugyot on bein hashemashot,
- Lists the topics to be covered in the remainder of the monograph,
- Explicitly specifies some halakhic assumptions, and
- Provides basic background on some relevant astronomy.
The part of this section posted below contains a discussion of the major sugyot concentrating on how Rabbeinu Tam and the geonim differentiate between the sugyot and a fuller discussion of the sugya in Shabbat.
The primary sugyot - Shabbat 34b and Pesachim 94a and their interrelationship; the basic opinions of the geonim and Rabbeinu Tam and the some of the fundamental challenges each position must address.
The gemara in Pesachim discusses whether the period between alot hashachar and sunrise (and the equivalent period from sunset to tzait hakochavim) is the time to walk 4 or 5 mil. The gemara in Shabbat describes the period of bein hashemashot in a variety of ways to be analyzed below, but when measured in time, sets the interval at either the time to walk ⅔ or ¾ of a mil. Although it is the tanna Rabi Yehudah’s opinion quoted in both sugyot, the contradiction could be resolved, as some have suggested, by assuming that perhaps Rabi Yehudah changed positions. However, given the significant discrepancy, almost all commentators attempt to resolve the inconsistency by postulating that the two sugyot are addressing differing intervals.
To ground this introduction, assume that the interval between alot hashachar and sunrise is the time to walk four mil and that the time to walk each mil is 18 minutes. On a canonical day, around the spring and fall equinox, sunrise and sunset are at 6AM and 6PM respectively. Under these assumptions, according to the gemara in Pesachim, alot hashachar is 4*18 or 72 minutes before sunrise at 4:48AM and tzait hakochavim is 72 minutes after sunset, at 7:12PM.
The conceptual approach of Rabbeinu Tam posits that the endpoints of the sugyot in Shabbat34b and Pesachim 94a are identical. The end of Shabbat, at the end of the bein hashemashot period in masekhet Shabbat, is at 7:12PM, 72 minutes after sunset. However, the beginnings of the periods differ between the sugyot. The gemara in Shabbat refers not to sunset, as we commonly refer to it, but a secondary sunset that occurs later, not just when the sun is no longer visible but rather when almost all of its light is longer visible as well. That occurs at 6:58 and 30 seconds, almost an hour after what we colloquially call sunset and at that time bein hashemashot begins. Until that time, the day continues, and on Friday night, work is permitted.
The geonim take exactly the opposite position. In their formulation, the beginning points of the two sugyot are (almost and according to many commentators exactly) identical. Thus, the period of bein hashemashot begins at sunset. Shortly thereafter, at 6:13 and 30 seconds, Shabbat ends on Saturday night and work is again permitted. At 7:12PM, after the full 4 mil period of the sugya in Pesachim, all the stars appear, not just the three medium stars that signify the end of Shabbat.
Practice has at various times and for a variety of reasons somewhat softened both positions. Nevertheless, it is critical to recognize that the disagreement is significant. In many European communities, Jews who followed Rabbeinu Tam worked well after what we refer to as sunset on Friday evening; many Jews living in Israel and points closer to the equator followed the geonim and ended Shabbat within 30 minutes after sunset. At least in terms of halakhic theory, a period of approximately an hour, defining both the beginning and end of Shabbat, is in dispute.
The approach of the geonim separates the tzait hakochavim discussed in masekhet Shabbat, approximately the time at which three stars appear, from the tzait hakochavim of masekhet Pesachim, the time that all the stars appear. It is therefore intuitively obvious why the period from alot hashachar to sunrise is identical in length to the period from sunset and the appearance of all the stars. All the stars appear when no remaining light from the sun impair their visibility; equivalently alot hashachar is coincident with the first rays of light in the morning. In the morning, as more light from the sun becomes visible, the number of stars that remain visible decreases; in the evening, the reverse occurs and as illumination from the sun disappears completely, all the stars (that can possibly be seen) become visible. An equal length of time between sunset and when light (however defined) disappears and again between the time light reappears and sunrise is the consequence of this symmetric definition. However, the approach of Rabbeinu Tam, that equates the tzait hakochavim of masekhet Shabbat and the tzait hakochavim of masekhet Pesachim, must deal with this issue of asymmetry. How can one equate
v the length of time between alot hashachar, when (almost) all of the stars are still visible and sunrise, with
v the length of time from sunset to the appearance of only three stars?
While the approaches of the geonim and Rabbeinu Tam are cited almost exclusively, it is critical to realize that other points of view are possible. The gemara in Pesachim defines a longer interval of 72 minutes. As normally assumed, the very short interval of approximately a quarter of an hour that is defined by the gemara in Shabbat is mapped by the geonim to the beginning of the 72 minute interval, while according to Rabbeinu Tam, it maps to the end of the 72 minute interval. However, at least conceptually, the interval in masekhet Shabbat may begin somewhere between the points suggested by the geonim and Rabbeinu Tam, beginning after sunset at 6:00PM but before 6:58PM and 30 seconds. We raise this, not just as a theoretical possibility. Rather, as will become evident, the strongest arguments that have been adduced in favor of the position of Rabbeinu Tam within the classical literature as well as those to be developed in this monograph, imply that the day continues after sunset delaying the start to the bein hashemashot period until sometime after sunset. Conversely, the greatest challenges to the opinion of Rabbeinu Tam result from his position that the end to the bein hashemashot period extends as late as the time to walk 4 mil after sunset. First, this requires that the astronomic conditions described in masekhet Shabbat defining the end of bein hashemashot are observable at the time it takes to walk four mil after sunset. Second, those conditions must be parallel to conditions that exist around alot hashachar. Neither is apparently true.
Therefore, for purpose of analysis, we will separate the argument between the geonim and Rabbeinu Tam into two parts:
v When does the period of bein hashemashot begin?
v When does the period of bein hashemashot end?
The length of the period (in the Middle East) is, according to all opinions, (perhaps as most currently assume only around the spring equinox) the time to walk ⅔ or ¾ of a mil.
The text of the gemara in Shabbat contains three sections that must be analyzed carefully. Unlike Rabi Yosi, who considers bein hashemashot as occurring instantaneously, during the blink of an eye, Rabi Yehudah, in a statement that the gemara rules as normative for the beginning of Shabbat, defines an interval of bein hashemashot in terms of the appearance of the horizon. Rabi Yehudah’s precise wording is challenged by the gemara as being inconsistent. It contains three initial phrases:
v one referring to the setting sun,
v one referring to the sun’s illumination and
v one referring to the darkening horizon
and then a fourth phrase that defines the end of the bein hashemashot period. Rabbah, who approximates the interval of bein hashemashot as the time required to walk ¾ of a mil, assumes all three initial phrases apply to the bein hashemashot interval. The difficulty with this approach is the repetitive nature of the description, requiring Rabbah to explain that the phrases refer to the beginning and two intermediate points within the bein hashemashot interval. The need for describing two intermediate points is somewhat forced. On the other hand, R. Yosef adds a word – “daytime” – and he explains the initial phrases describe an interval after sunset that is still daytime, the second phrase describes the appearance of the (western) horizon just prior to the beginning of bein hashemashot and the third phrase describes a point during bein hashemashot. The difficulty with this approach is that the word “daytime” has to be added and must be assumed to be understood implicitly in the original formulation of Rabi Yehudah. This slightly shorter interval for bein hashemashot aligns with R. Yosef’s position that maintains the length of the interval is the time to walk only ⅔ of a mil interval, 1/12 of a mil shorter than Rabbah’s ¾ of a mil interval.
The gemara then records a discussion between Abaye and Ravah debating whether one should be looking at the western or eastern horizon; this passage is critical and must be explained according to the interpretations of both the geonim and Rabbeinu Tam.
Both Rabbah and R. Yosef attribute their opinion of the position of the tanna Rabi Yehudah to the opinion of the amora R. Yehudah in the name of Shmuel. Later, the gemara, without any suggestion of disagreement, quotes another statement of R. Yehudah in the name of Shmuel concerning 1, 2 and 3 stars. We address alternative interpretations of this statement and its relationship to the remainder of the sugya in a subsequent section.
With respect to the argument of Rabbeinu Tam and the geonim, a number of other sugyot are referenced in support of the geonim. One that is representative is the gemara in Zevachim that asserts that (sacrificial) blood is disqualified at sunset, seemingly in opposition to the opinion of Rabbeinu Tam. In such instances, one of two arguments can be invoked:
v The halakha that specifies precisely sunset only applies to a unique situation, as in this case of korbanot.
v The language is less clear than one might assume, and sunset refers not its precise astronomical definition but to some later point later when the sun’s impact is diminished.
As these sugyot have been extensively quoted and debated, it is unnecessary, except on rare occasion, to review them in this monograph. Instead, we concentrate on analysis of the basic sugyot mentioned.
One additional source that bears on these two sugyot is the lengthy discussion in the yerushalmi in Berachot 2b that considers the verse in Nehemiah 4:15:
v Veanachnu osim bemelacha…..mei alot hashachar ad tzait hakochavim
as defining the daytime period from alot hashachar until the appearance of three stars. Rabbeinu Tam is entirely consistent with this approach; in fact, it provides an explicit sugya that links the appearance of three stars in the evening as the endpoint corresponding to alot hashachar in the morning. How the asymmetry is to be dealt with is very different, however, from the overall problem of asymmetry that Rabbeinu Tam must address. Unlike the gemara in Pesachim, nothing in the yerushalmi, and the other texts that quote the verse in Nehemiah, (even remotely) implies symmetry. Those sugyot that quote the verse in Nehemiah appear to support a definition of the daytime period that is asymmetric relative to both chatzot and the time from sunrise/sunset. While this asymmetry may be surprising, it is not, in and of itself, an issue. On the other hand, the gemara in Pesachim explicitly introduces symmetry by asserting equal intervals bracketing sunrise and sunset, each of the same duration, the time to walk either 4 or 5 mil. The opinion of Rabbeinu Tam further extends the symmetry of the gemara in Pesachim by equating the endpoints of the sugyot in Shabbat and Pesachim. By implication, Rabbeinu Tam equates the endpoint in the yerushalmi as well. For the geonim, however, this gemara in Shabbat (and the yerushalmi in Berachot) is addressing a different endpoint than the (later endpoint referenced in the) gemara in Pesachim.
Two approaches to this later point, tzait kol hakochavim, when all the stars appear are possible. According to the Gaon, the gemara in Pesachim is theoretical; the evening equivalent of alot hashachar, 4 mil after sunset, when all the stars appear, is ascribed no examples of halakhic significance. Alternatively, one can assume that the while the appearance of three stars approximates or defines the end of each day of the week, the daytime period extends beyond that point, until all illumination from the sun has disappeared. That later point may apply to specific areas of halakha, where only the daytime period as opposed to the specific day of the week is relevant, a theoretical possibility addressed in section 9.
The geonim must deal with a number of questions. First, since it is universally acknowledged that the day starts with alot hashachar, should it not correspondingly end at an equivalent point after sunset i.e. tzait (kol) hakochavim? This question is easily addressed. Symmetry is required only according to the approach of Rabbeinu Tam; for the geonim, given the two distinct definitions of day, the day of the week versus the daytime period, symmetry need not be expected. Second, and this question is fundamental, after an interval of the time required to walk ¾ of a mil after sunset, in the Middle East, three medium stars are barely visible, if at all.
Rabbeinu Tam must address a number of fundamental questions. First, In the Middle East three medium stars appear well before the time required to walk 4 mil after sunset. Second, the description of the horizon around the end of Shabbat as well as the debate between Abaye and Ravah all appear to support the opinion of the geonim. Third, and most fundamentally, how could the time of appearance of three stars and the time of alot hashachar be identically separated (as specified by the gemara in Pesachim) from sunset and sunrise respectively? The analog to three stars becoming visible is not alot hashachar when (almost) all stars are still visible. This question and the debate between Ravah and Abaye in Shabbat, present major challenges to the opinion of Rabbeinu Tam.
We return to these two sugyot as we analyze the primary halakhic categories below, attempting to formulate responses to the above questions throughout the monograph.
 We often will use the phrase the time to walk X mil. Undoubtedly, a person walking for an entire day will cover fewer mil per hour than one who walks for a shorter period. That adjustment is not how this system of specifying time periods is defined. Instead, if one walked 32 mil in a daytime period of 12 hours, the time to walk a single mil is 22.5 minutes, 12*60/32, despite the fact that walking only one mil takes significantly less than 22.5 minutes. Similarly, the number of mil covered in 90 minutes is 90/720, (1/8) * 32 mil = 4 mil.
 We use “Rabi” to denote a tanna and “R.” for an amora.
 It is Rabi Yehudah quoted explicitly in Pesachim specifying that bein hashemashot is 1/10 of a day; in Shabbat there is a dispute between Rabbah and R. Yosef concerning the length of Rabi Yehudah’s bein hashemashot period.
 Both the length of time to walk a mil and the number of mil in the twilight interval are areas we cover in detail in Sections 2 and 3. Choosing either 22.5 minutes or the more typical 18 minutes as the time to walk a mil, is not consequential to this introduction. To demonstrate that, the post chose the more typical 18 minutes, while the pdf chose 22.5. As a result we also use the more typical 72 (4*18) minutes in the post versus the 90 (4*22.5) minutes that is used for illustration in the attached pdf.
 A careful reading of Rabbeinu Tam (and other poskim) might align 6PM with the sun beginning to go below the horizon, a few minutes prior to the sun having gone completely below the horizon, which is our usual definition of sunset. Note that this makes the day at the equinox exactly 12 hour, while chatzot is a few minutes early. We disregard this and other minute differences.
 The potential halakhic consequences, if any, of the appearance all the stars is discussed in section 9.
 This issue is most fundamental. As we plan to demonstrate, it is impossible to address the issue fully without radically changing some fundamental element of Rabbeinu Tam’s conceptual opinion and/or how it is practiced. R. Soloveitchik did exactly that both in (personal) practice and in the theory developed in his yarzeit shiur. Absent so major a change in Rabbeinu Tam’s approach, attempts to deal with this issue will introduce other complications. One representative issue is that chatzot, when defined as the midpoint between alot hashachar and the appearance of only three stars, will occur too early – not at the precise point when the sun is directly overhead. Many, as illustrated by Benish, simply disregard the issue or assume that chatzot need not be precise. Benish cites many examples of calendars from major Jewish communities, ostensibly endorsed by their rabbinic leaders would often calculate chatzot assuming the appearance of three stars and dawn are symmetric endpoints. There are reasonable options to address this issue; most often, however, the issue was not addressed adequately. In a somewhat similar fashion, R. Feinstein’s tshuva allowing early morning prayers 90 minutes before sunrise (Igrot Moshe OC 4: 6) and his tshuva on nighttime zemanim (Igrot Moshe OC 4:62,) are not cross-referenced. This basic issue is addressed in multiple contexts in this monograph.
 Alternatively, at 7:13PM and 7.5 seconds, if the interval is 90 minutes ending at 7: 30PM. We do not deal with the isolated opinion of the R. Eliezer miMitz, the author of the Yeraim that posits that the interval in masekhet Shabbat begins approximately 15 minutes before sunset.
 Perhaps, as is subsequently debated by Ravah and Abaye, the two intermediate points are separately describing conditions in the eastern and western sky.
 This debate between Abaye and Ravah is covered in section 7.
 This statement is analyzed carefully in Section 5.
 Tosfot’s answers in Zevachim 56a s.v. menayin ledam and the discussion in Minkhat Kohen, ma’amar rishon chapters 4, 10, 11, 12 and 14 are prototypical.
 Minkhat Kohen organizes and comprehensively covers the major sugyot on this issue.
 Discussion of the verse occurs in more abridged form in the bavli as well – Megillah 20a and Berachot 2b.
 OC 261.
 R. Soloveitchik in his yartzeit shiur previously referenced, developed and advanced an argument that clearly affirmed this type of symmetry as he reformulated the opinion of Rabbeinu Tam. However, as will be discussed, the way Rabbeinu Tam was practiced, did not always comply with this principle of symmetry.
 It is somewhat surprising that this question was not broadly discussed until the 20th century. Perhaps, consistent with our central thesis, the Sabbath did not begin precisely at sunset as most now assume, even according to the geonim.
 This is covered in more detail in section 7.
 As will be discussed further in detail, only approaches like those of R. Soloveitchik fully tackle the fundamental questions raised by symmetry with respect to Rabbeinu Tam’s position.