Skip to main content

Full text of "Midrash and Mishnah a Study in the Early History of the Halakah: I"

See other formats


Early Journal Content on JSTOR, Free to Anyone in the World 

This article is one of nearly 500,000 scholarly works digitized and made freely available to everyone in 
the world by JSTOR. 

Known as the Early Journal Content, this set of works include research articles, news, letters, and other 
writings published in more than 200 of the oldest leading academic journals. The works date from the 
mid-seventeenth to the early twentieth centuries. 

We encourage people to read and share the Early Journal Content openly and to tell others that this 
resource exists. People may post this content online or redistribute in any way for non-commercial 

Read more about Early Journal Content at 
journal-content . 

JSTOR is a digital library of academic journals, books, and primary source objects. JSTOR helps people 
discover, use, and build upon a wide range of content through a powerful research and teaching 
platform, and preserves this content for future generations. JSTOR is part of ITHAKA, a not-for-profit 
organization that also includes Ithaka S+R and Portico. For more information about JSTOR, please 



By Jacob Z. Lauterbach, Hebrew Union College. 

THE teachings of the Halakah, as preserved to us in 
the tannaitic literature, have been given by teacher to 
disciple and transmitted from generation to generation in 
two different forms, namely, Midrash and Mishnah. The 
one, Midrash, shortened from 'Midrash Torah', 1 represents 
the Halakah as an interpretation and exposition of the 
Torah. It teaches the Halakah together with its scriptural 
proof, that is, in connexion with the passage from the 
Pentateuch, on which it is based or from which it can 
be derived, thus forming a halakic commentary to the 
written law contained in the Pentateuch. This form is 
especially used in our halakic Midrashim, Sifra, Sifre, and 
Mekilta, but it is also found in some parts of the collections 

1 The term BH1D from Em 'to search, inquire, investigate', means 
' research, inquiry ', and mm BmD accordingly means an inquiry into 
the meaning of the Torah, an exposition of all laws and decisions which 
can be discovered in the words of the Torah. In this sense the term 
'Midrash Torah' is used in the Talmud (b. Kiddushin 49b) where it 
designates the halakic interpretation or exposition of the Torah. As we 
now have many Midrashim to the Torah of a haggadic character, the term 
Midrash Torah would be too indefinite to designate an halakic exposition 
of the Torah. A haggadic exposition of the Torah would also be a Midrash 
Torah. The more specific term Midrash Halakah is therefore now used 
to designate a halakic interpretation of the Torah. See the writer's 
article ' Midrash Halakah ' in the Jewish Encyclopaedia, VIII, pp. 569-73. 

VOL. V. 5°3 L 1 


of our Mishnah and Tosefta, as well as in many so-called 
Midrash-Baraitot scattered in both the Palestinian and the 
Babylonian Talmud. The other form, the Mishnah, repre- 
sents the Halakah as an independent work, giving its dicta 
as such, without any scriptural proof, and teaching them 
independently of and not connected with the words of the 
written law. For this reason the Mishnah is also designated 
as ' Halakah ' or in the plural ' Halakot ', that is, merely 
rules or decisions. This form is especially used in our 
collections of the Mishnah and the Tosefta, but it is also 
found in many Baraitot scattered in the Talmud and in 
some parts of our halakic Midrashim. 2 (See D. Hoffmann, 
Zur Einleittmg in die halachischen Midraschim, Berlin, 
1887, p. 3.) 

Of these two forms of teaching the Halakah, the 
Midrash is the older and the Mishnah the later. The 
Midrash was the original form, and was used in the earliest 
times, in the very beginnings of the Halakah. This is 
quite self-evident, as the Midrash was in reality the origin 
of the Halakah. The dicta of the Halakah had their 
source in the Midrash Torah, i.e. an inquiry into the full 
meaning of the written law from which alone the earliest 
Halakah derived its authority. 

The returned Babylonian exiles, constituting the new 
Jewish community, reorganized by Ezra and Nehemiah, 
accepted the written Torah, so to speak, as their constitu- 
tion. They entered into a covenant by oath, to keep and 
follow the laws of Moses as contained in the book read 

2 As the difference is only in form, it is not surprising to find that very 
many of the Halakot are cast in both forms. Very often the same Halakot 
which are found in the halakic Midrashim together with their scriptural 
proofs are also found in the Mishnah and Tosefta without scriptural proofs, 
as independent Halakot. 


to them by Ezra (Neh. 8 and 10. 30). The Book of the 
Law, therefore, as read and interpreted by Ezra, was for 
them the only authority they were bound to follow. 
Whatever was not given in the book, they were not bound 
to accept. All the religious practices and the time- 
honoured customs and even the traditional laws, if there 
were such, had to receive the sanction of the written Law 
in order to be absolutely binding upon the people. This 
means, that the practices, customs, &c, had to be recog- 
nized as implied in the written Law or contained in its fuller 
meaning. The teachers, therefore, interpreted the written 
Law so as to include in it or derive from it all those 
customs and practices. Thus, the teachings of the Halakah 
(for all such rules, customs, practices, and traditional laws 
constituted the Halakah) had to be represented as an 
interpretation or an exposition of the written Law. This, 
as we have seen above, means, to be given in Midrash-form. 
It is expressly stated of Ezra that he explained and 
interpreted the Torah to the people, and that he set his 
heart to search (WTO) the meaning of the Law, to 
interpret it, and to teach in Israel statutes and judgements 
(Ezra 7. 10). We learn from this, that Ezra taught only 
the Book of the Law with such interpretations as he could 
give to it. His successors, the Soferim, who were the 
earliest teachers of the Halakah, did the same. They gave 
all their teachings merely as interpretations to the Book of 
the Law. Indeed, the very name Soferim was given to 
them because it characterized their manner of teaching. 
This name D'lSlD is derived from 1SD 'the Book'. It means 
' Bookmen', and it designated a class of people who occupied 
themselves with the Book of the Law, who interpreted it 
and who based all their teachings upon this book exclu- 

L 1 2 


sively (Frankel, Hodegetica in Mischnam, p. 3, and Weiss, 
Dor, I, p. 47). 

For a long period this Midrash-form was the only form 
used in teaching the Halakah. This is confirmed by reliable 
traditions reported to us in Rabbinic literature. One such 
report is contained in the following passage in the Pal. 
Talmud (Moed katan III, 7, 83") : 

->dk mm Tiyi nvbn rrw f>a hdn irprn ?D3n Tof>n wk 
.httivi l^ax may bx rowtna i»x n*n *nrt w 't n^ 

' Who is to be considered a scholar ? Hezekiah says, 
One who has studied the Halakot as an addition to and in 
connexion with the Torah. 3 Said to him R. Jose, What 
you say was [correct] in former times, but in our day, even 
[if one has studied merely detached] Halakot, [he is to be 
regarded as a scholar].' Here it is plainly stated that in 
earlier times (rDlEWO) the only form of teaching Halakot 

3 The term Tlj?1 means 'addition', as, for instance, in the phrase: 
&fnp? "liyi miiT 1 ! ' Is it necessary to mention the custom in Judea as an 
addition to the law indicated in the Scriptures?' (b. Kiddushin 6a). It is 
also found in the plural form, nVTUTl 'additions' (b. Erubin 83a). The 
expression miD 11$)1 here means, therefore, as an addition to the Torah, 
i.e. to teach the Halakot not independently but as additions to the passages 
in the Torah from which they are derived. In almos' the same sense it is 
also interpreted by the commentator Pne Mosheh, adloc. 

It should also be noticed that in b. Kiddushin 49 a Hezekiah says that 
to' be called a student (flJlC?) it is enough if one has studied merely detached 
Halakot. This, however, does not contradict his saying in our passage in 
the p. Talmud. For d3n T l D?n is a scholar of a higher degree of learning. 
From b. Megillah 26 b it is evident that the student called nJV<5> is not as 
advanced as the scholar called D3n T'DPfl . To be considered a scholar, 
such as is designated by the name D3n T'DPf) , Hezekiah tells us, one must 
study the Halakot in the Midrash-form. For even after the Mishnah-form 
had become popular, the Midrash was considered the proper form to be 
used by advanced scholars. See Guttmann, Zur Einhiiung in die Halakah, 
Budapest, 1909, p. 20. 


was as an addition to and in connexion with the written 
Law, that is to say, in the Midrash-form. In those days, 
therefore, one could not acquire a knowledge of the 
Halakah, i. e. become a scholar, except by learning the 
Midrash, for the very good reason that the halakic 
teachings were not imparted in any other form. 

Sherira Gaon who no doubt drew upon reliable sources 
likewise reports in his Epistle (Neubauer, M. J., ch. I, p. 15) 
that ' in the earlier period of the second temple, in the 
days of the earlier teachers, all the teachings of the Halakot 
were given in the manner in which they are found in our 
Sifra and Sifre ', that is, in the Midrash-form. 4 Modern 
scholars s have, accordingly, recognized it as an established 
historic fact that the Midrash was originally the exclusive 
form in which all teachings of the Halakah were given. 

Not only were those Halakot which were derived from 
some scriptural passage by means of interpretation taught 
in Midrash-form, that is to say in connexion with the 
passages which served as proof, but also such Halakot 
and teachings as were of purely traditional origin — rules, 
practices, and customs that had no scriptural basis at all 
were likewise taught in this manner. The latter were taught 
in conjunction with some scriptural passage with which 
they could in some manner be connected, or together with 
certain written laws to which they were related, either as 

* The passage in the letter of Sherira Gaon reads thus : HED1 SOBDt 

>:v enpon vnpym 'top:: xna^n vo-\ p'm iroy wpi ">um 
pr6 ■on inn kityik inn dis& ••kdp pm movn. They taught 

'them', i. e. the Halakot, only in the form used in our Sifra and Sifre, 
i. e. Midrash. 

6 N. Krochmal in More Nebuke Ha-Zeman, porta XIII, Lemberg, 1851, 
pp. 166-7; Z. Frankel in Hodegetica in Misc/mam ;'We\ss,DorDorwe-Dorshow 
and Mabo la-Mechilla; Oppenheim, 'Toledot ha-Mishnah' in Beth Talmud, 
II ; D. Hoffmann, Die erste Mischnah, Berlin, 1882 ; and others. 


corollary or modification. (See D. Hoffmann, Die erste 
Mischnah, Berlin, 1882, pp. .5-7.) This procedure was 
necessary, because the only recognized authority was the 
written Book of the Law which the teachers used as their 
text-book in teaching. However, in teaching out of this 
text-book, they gave not only the meanings of words and 
the explanations of each written law, but also additional 
rules as well as modifications to some laws. All of this 
may be included in an exposition (Emo) of the Torah and 
could properly be taught in connexion with the text. 
Thus the Midrash-form could continue to be in exclusive 
use for teaching the Halakah, even after the latter, in the 
course of time, came to include traditional laws and 
customs, as well as new institutions and decrees issued and 
proclaimed by the teachers themselves in their capacity as 
religious authorities. 6 

The Mishnah-form, on the other hand, is of a much 

later date. It was introduced a long time after the 

Midrash-form 7 and was used side by side with it. At 

• Weiss, Mabo la-Mechitta, p. iv, remarks about the Soferim : bioStJ* 

ppvim mw D-iipe lent cnm dj ife ttnpcb dpitsi Deu-in. 

Although the instance mentioned by him as proof for his statement is not 
a teaching of the Soferim (see below, note 55), yet the statement as such 
is correct. The Soferim or those who only taught in the Midrash-form 
could include in their teachings altogether new laws and decrees, issued 
by themselves as religious authorities, by connecting them with the 
scriptural laws. Only we may assume that it rarely happened that they 
taught a traditional law or a decree of their own merely in connexion with 
some scriptural law. In most cases, the Soferim, who had charge of the 
text of the books of the law, could manage to indicate in the text itself, 
by means of certain signs and slight alterations, any traditional custom 
or decree of their own. Thus, these same decrees could be taught as 
interpretations of the written law. See N. Krochmal, op, cit., p. 167. 
Compare also below, notes 36 and 37. 

T Georg Aicher {Das AlU Testament in der Mischnah, Fr.-i.-Br., 1906, 
pp. 165 ft".) stands alone in the assumption that the Mishnah is older than 


no time did the Mishnah-form become the exclusive 
method for teaching the Halakah, because the Midrash 
never ceased to be in use. 8 At just what date this 
Mishnah-form was introduced, that is to say, just when 
the teachers of the Halakah began, for the first time, to 
teach Halakot independently of the written law, has, to my 
knowledge, not yet been ascertained. Sherira Gaon who, 
as we have seen, informs us that at some period in earlier 
times the Midrash-form was the only one in use, does not 
state exactly how long that period lasted, and does not 
mention when the Mishnah-form was introduced. Neither 
is there any other gaonic report to tell us when this 
happened. 9 Hoffmann (pp. cit., pp. 12-13) states that, 
according to the views held by the Geonim, the Mishnah- 
form was first introduced in the days of Hillel and 
Shammai, but he fails to bring proof for this statement. 
To my knowledge, there is no foundation in gaonic 
literature for the views ascribed by Hoffmann to the 
Geonim. Hoffmann bases his theory on the spurious 

the Midrash. This cannot be maintained. His statement (p. 64) that 
' the appearance of scriptural proof in connexion with the Halakah was due 
to the radical changes effected by the catastrophe of the year 70 ', hardly 
needs any refutation. The many Halakot in the Midrash form given by 
teachers in the time of the Temple as well as the disputes between the 
Sadducees and Pharisees, hinging upon different interpretations of scriptural 
passages as bases for their respective Halakah, ought to have shown Aicher 
to what extent Midrash was used before the year 70. 

8 We must emphasize this fact against the theory advanced by Weiss 
and Oppenheim and also by Jacob Bassfreund in his Zur Redaction der 
Mischnah (Trier, 1908, pp. 19-24), that there was a time when the Midrash- 
form was altogether abandoned, and the teachings of the Halakah given 
exclusively in Mishnah-form. We shall see that this theory is untenable 
(below, notes 15, 23, and 53). 

9 The account given in the letter of Sherira stops very abruptly. See 
the discussion at the end of this essay. 


responsum found in Shaare Teshubah, No. ao, and ascribed 
to Hai Gaon, 10 in which the following passage is found : 

i»3 two niD niND w vn jprn V?n iy «'n two nwo jn 
r^iyn pddjw oyoru "]^K1 ^n jdi wdi rwrh tfl'n djjw 
anno nw t6a *kde>i 9?nn upn k$>i min i>c rrrna ne^m 


' Know, that from the days of Moses our Teacher until 
Hillel the Elder, there were six hundred orders of Mishnah 
just as God gave them to Moses on Sinai. However, from 
the time of Hillel on the world became impoverished, and 
the glory of the Law was diminished, so that, beginning 
with Hillel and Shammai, they arranged only six orders.' 
It is evident that this responsum cannot be taken to 
represent a reliable gaonic tradition, as it is apparently 
based on the haggadic passage in Hagigah 14 a, and is 
accordingly of merely legendary character. Aside from 
this, the passage does not say what Hoffmann has read 
into it. It does not even deal with the origin of the 
Mishnah-form. If anything, we can see from this respon- 
sum that its author, quite to the contrary, assumed that 
the Mishnah-form was very old, and that it was given to 
Moses on Sinai. 11 He deals merely with the origin of six 

10 This responsum had been added by some later hand to the responsa 
of Hai Gaon, but does not belong to the Gaon. Comp. Harkavy, Studien 
und Mitteilungen, IV, p. xiv. The fact that this report is repeated in 
Seder Tannaim we-Amoraim, (Breslau, 1871, p. 29) and in Sefer Hakanah, 
p. 81 b, and in S. Chinon's Sefer Kritot (Book Yemot Olam, Amsterdam 
1709, p. 20 a) does not in the least alter its legendary character and cannot 
make it more reliable, for the authors of all these works drew from one and 
the same source. This source cannot be of a more reliable character than 
the Midrash Abkir, from which the Yalkut (Genesis, sec. 42) quotes the 
statement that Methuselah studied 900 orders of Mishnah, pHJ? TVTWD 

roe>» mo niKD 'd rw rpm n\n -mm. 

11 The belief that the Mishnah was given to Moses on Sinai is repeatedly 


orders of Mishnah which he assumed to have been extant 
in the days of Hillel and Shammai. These six orders 
were in his opinion but a poor small remnant of the six 
hundred orders which Moses received from God on Sinai 
and which were extant till the days of Hillel when the 
world became impoverished and the glory of the Torah 
diminished. Hoffmann arrives at his interpretation of this 
responsum by arbitrarily giving two different meanings to 
one and the same term used by the author twice in one 
sentence. He states (p. 13) that when the Gaon speaks of 
the ' six hundred orders of Mishnah ', he is using the term 
' Mishnah ' in a broad sense to designate traditional law in 
the Midrash-form and not in the Mishnah-form, but when 
the Gaon speaks of the reduced ' six orders ' extant in the 
days of Hillel and Shammai, he uses the term ' Mishnah ' 
in a narrow sense to designate only independent Halakot 
in the Mishnah-form. This distinction is extremely arbitrary. 
Furthermore, when Hoffmann concludes his argument with 
the remark (ibid., p. 13) that ' No doubt the six orders of 
Mishnah introduced in the days of Hillel and Shammai 
were, like our present Mishnah, composed in the form of 
independent Halakah, and by this new form were distin- 
guished from the earlier form of teaching', he no longer 

expressed in the Haggadah. See b. Berakot 5 a and p. Hagigah I, 8,, 76 d. 
In the Pirke de-Rabbi Eliezer, ch. xlvi, it is said that during the forty days 
which Moses spent on the mountain, receiving the Law, he studied the 
Scriptures (N*TpE>) in the daytime and Mishnah at night. In Pesikta 
Rabbati V (Friedmann, p. 14 b) it is said that Moses wished to have the 
Mishnah written, but God told him that in order to distinguish Israel from 
other nations it was better that the Mishnah should be given to Israel 
orally, so that the other nations should not be able to claim it for themselves. 
See also Tanhuma, Ki-Tissa (Buber, pp. 58b and 59a), and p. Hagigah, /. c. 
The author of our responsum had as his authority such haggadic sayings 
when he spoke of the Mishnah which God gave to Moses on Sinai. 


gives the views of the author of the responsum, but his 
own. And these views are absolutely wrong. 12 

Thus we see that there is no mention in gaonic literature, 13 
of the time when this innovation in the form of teaching 
the Halakah took place. Neither is there any report in 
talmudic 14 or gaonic sources about the cause of this 
innovation. We are not told why it was necessary or 
desirable to introduce a new form of teaching Halakah 
alongside of the older Midrash-form. 

Modern scholars have attempted to answer these 
questions; both to fix the date and to give the reasons 
for this innovation in the method of teaching. However, 
the various theories advanced by these scholars are all 
unsatisfactory. They are the result of mere guess-work — 
without solid proof or valid foundation. It will be shown 

12 There is no doubt that at the time of Hillel and Shammai there were no 
Mishnah-collections like our Mishnah. The responsum in Shaare Teshubah, 
§ 187, which tells us that when a certain Gaon died they found that he had 
the six orders of the Mishnah of the days of Hillel and Shammai, which had 
been hidden away, is spurious and legendary. See S. D. Luzzatto, Beth 
ha-Ozar, pp. 55b-s6a. Although there were in the times of Hillel and 
Shammai collections of Halakot composed in Mishnah-form, this form was 
not new to them and could not be the characteristic which distinguished 
them from the form of teaching used before. For, as we shall see, there 
had been even before Hillel and Shammai collections of independent 
Halakot in the Mishnah-form. And if Hillel himself composed a Mishnah- 
collection, he did not arrange it in order, and did not divide it into tractates 
as Pineles (Darkah shel Torak, pp. 8-9) and Bassfreund (Zur Redaction der 
Mischnah, p. 25) assume. The arguments brought forward by the latter to 
prove that Hillel's Mishnah-collection was arranged and divided into tractates 
are not convincing. 

13 On Saadya's opinion see further below. 

14 There is, however, as we shall see in the course of this essay, a report 
in the Talmud stating until when the Midrash-form was in exclusive use. 
This talmudic report has been overlooked or else not correctly understood, 
for not one of the scholars dealing with the problem of fixing the date of 
the beginning of the Mishnah-form has referred to it. 


that some are based upon inaccurate reasoning, and all of 
them are in contradiction to certain established historic 

We have already seen that the theory which Hoffmann 
ascribes to the Geonim has no foundation in gaonic 
literature and that it is altogether Hoffmann's theory. 
But, no matter whose it is, the theory itself cannot be 
maintained. 15 In the first place, there were Mishnah- 
collections before the time of Hillel and Shammai, as 
Rosenthal has proved {Ueber den Zusammenhang der 
Mischnah, Erster Teil, ate Aufl., Strassburg, 1909). In the 
second place, the introduction of a new form necessarily 
precedes any collection of Halakot composed in this new 
form. It must be quite plain that there were individual, 
detached Halakot taught in the Mishnah-form (and not in 
the Midrash-form) before any collection of such detached 
Halakot could be made. Accordingly, if we assume with 
Rosenthal (pp. cit., p. 111) that a collection of such inde- 
pendent Halakot in the Mishnah-form was already arranged 
in the time of Simeon ben Shetah, we have to go still farther 
back in fixing the time when the teachers first began to 
separate the Halakah from its scriptural proof and teach it 
independently, as Mishnah. This would bring us to about 
one hundred years before the time of Hillel and Shammai. 
Not only is this theory of Hoffmann wrong in respect to 

15 Compare also Bassfreund (op. cit., pp. 18 ff.) who likewise seeks to 
refute Hoffmann's theory. Some of Bassfreund's arguments, however, are 
not sound. He is altogether wrong in assuming that for a long time before 
Hillel the Mishnah was the exclusive form used in the teaching of the 
Halakah, and that Hillel was the first to reintroduce the Midrash-form. 
He confuses the development of the Midrash methods which were furthered 
by Hillel with the use of the Midrash-form which had no need of being 
introduced by Hillel since it was never abandoned (see above, note 8, and 
below, note 32). 


the date given for the introduction of the Mishnah-form, 
but it is also unsatisfactory in regard to the cause of this 

According to this theory, the Mishnah-form was intro- 
duced in order to assist the memory in mastering the 
contents of the traditional law. 16 However, it is difficult 
to see how the teachers could have considered the new 
form of greater aid to the memory than the old form. 
This new form is on the contrary quite apt to make it 
more difficult for the memory. It seems to us that it is 
less of a task for the memory to retain Halakot taught 
in the Midrash-form. The written Law, being the text- 
book, each passage in it, as it is being read, helps, by 
mental association, to recall all the halakic teachings 
based upon it. On the other hand, it is much harder 
to remember detached Halakot given in an independent 
form, especially when they are not arranged systematically 
or topically but merely grouped together. This, we must 
keep in mind, was actually the mode of arrangement used 
in the earlier Mishnah collections. 17 

Hoffmann himself must have felt that this theory was 
not satisfactory, for later in his book he advances another 

10 The same reason is also given by Frankel and Weiss. They all seein 
to have been influenced by the haggadic sayings found in the Talmud, 
sayings which exaggerate the number of Halakot known to former 

17 Hoffmann makes the mistake of assuming {op. cit., pp. 13, 15, and 48) 
that simultaneously with the separation of the Halakot from their scriptural 
basis came the grouping of such detached Halakot into orders and treatises, 
as we have them. But this is absolutely wrong. The earlier Mishnah went 
through many different forms of grouping before it was finally arranged 
according to subjects and divided into treatises and orders. See the writer's 
article in the Jewish Encyclopaedia, VII, p. 611. The opinions expressed 
by the writer there on page 610 (following Hoffmann) are hereby retracted. 


and altogether different theory {pp. cit., p. 48). According 
to this second theory, the innovation was not made for the 
purpose of aiding the memory, and was not made in the 
days of Hillel and Shammai. Here Hoffmann assumes 
that the Mishnah-form was first introduced in the days 
of the later disciples of Hillel and Shammai. The purpose 
of the innovation, he explains, was to maintain the unity of 
the Halakah by minimizing the differences of opinion and 
eliminating the disputes about the halakic teachings which 
arose among these very disciples of Hillel and Shammai. 
These disputes, Hoffmann tells us, were in many cases only 
formal, namely, concerning the underlying Midrash or the 
scriptural proof for the halakic teaching. The traditional 
Halakah, as such, was agreed upon by all the teachers. 
That is to say, there was no dispute about the transmitted 
rules and decisions which all the teachers received alike. 
The teachers, however, often did disagree as to the 
scriptural passages and their interpretations whereon these 
received halakic decisions were based. One teacher would 
derive a certain Halakah by interpreting a given passage 
in a certain manner. Another teacher would deduce the 
same Halakah from another passage, or even from the 
same passage but by means of another interpretation. 
Thus, as long as the Halakah was taught only in Midrash- 
form there existed many differences of opinion between 
the teachers, not in regard to the halakic decisions or 
rules in themselves but in regard to their midrashic proof 
and support. The teachers of those days who were very 
anxious to maintain harmony among themselves and 
unanimity in their teachings therefore decided to separate 
the Halakah from the Midrash and to teach it inde- 
pendently of the scriptural proof or support. In other 


words, they introduced the Mishnah-form— the Halakah 
as an independent branch of learning. By this innovation 
all the differences of opinion and disputes about the 
midrashic proof necessarily disappeared. Thus uniformity 
was restored in teaching the Halakah, and harmony was 
established among the teachers. 

This second theory of Hoffmann is even less tenable 
than the first. In the first place, it fixes the date for the 
introduction of the Mishnah even later than the first theory. 
Consequently, in this respect it is refuted by the same 
arguments that were brought against the first theory. 
We have seen above that there were Halakot in Mishnah- 
form, even collections of such Halakot, at a much earlier 
date. Furthermore, the explanation of the cause for the 
innovation put forth in this theory presents a palpable 
error in reasoning. It presupposes that the decisions of 
the Halakah, as such, were older than their midrashic 
connexion with the scriptures, and that at some earlier 
time they had been transmitted independently of scriptural 
proofs. For this reason the teachers could well be 
unanimous in accepting the Halakah and yet find cause for 
dispute as to methods of proving certain halakic decisions 
from the scripture by means of the Midrash. But this 
means nothing else than that there were some Mishnahs. 
that is, independent Halakot before the disputes about the 
scriptural proofs caused their separation from the Midrash. 
This line of reasoning contradicts itself. It sets out to find 
the cause for the first introduction of the Mishnah-form, 
but assumes that before this introduction some Halakot 
had already been transmitted in Mishnah-form. In other 
words, this so-called first introduction was really not a first 


If they had taught only in Midrash-form, the alleged 
evil results which the Mishnah-form, according to Hoffmann, 
was to remedy could never have arisen. It would have 
been impossible for the teachers to agree upon a halakic 
decision, and at the same time to disagree about its 
scriptural proof. Since every teacher received each 
Halakah in the same Midrash-form, that is, as an in- 
terpretation of, or connected with, a certain scriptural 
passage, every one who remembered the decision must 
have remembered the form in which he received it, that 
is, the scriptural passage with which it was connected. 
It is very improbable that a teacher remembering the 
decision, but having forgotten the scriptural basis, would 
have supplied another scriptural proof therefor, and then 
disputed with his colleagues who remembered the right 
passage on which this Halakah was based. If he did 
forget the passage for which the Halakah was an inter- 
pretation, the mere mention of that passage by his colleagues 
must have brought it back to his memory. It is evident 
that there could be no universal acceptance of a Halakah 
together with disputes regarding its proofs, unless such a 
Halakah had been taught apart from its proof. This, 
however, was not done, as long as the Midrash-form was 
in exclusive use, that is, as long as the Halakah was merely 
taught as a commentary on the text of the Law. 18 

18 This would hold true even if we should believe in the genuineness 
of the so-called WDO TVtftb lTD^n, that is, that there had been given oral 
laws to Moses on Sinai and transmitted independently of the written law. 
For, as Hoffmann himself states {op. cit., p. 7), even all the traditional 
teachings were taught together with the scriptural laws and connected 
with them in the Midrash-form. All through the period of the Soferim, 
and according to Hoffmann till the time of the disciples of Hillel and 
Shammai, such traditional laws would somehow be connected with the 
Scriptures. The mental attitude of the teachers was not in the direction 


Quite as unsatisfactory is the theory advanced by 
Z. Frankel (Hodegetica in Mischnam, pp. 6, 7, and 10). 
According to this theory, the innovation of teaching de- 
tached Halakah in the Mishnah-form was made by the 
last group of Soferim. 19 This was done to overcome three 
difficulties which Frankel tells us existed in those days. 
In the first place, the halakic decisions based upon the 
individual passages had increased to such an extent that 
the task of studying and teaching them in the Midrash- 
form became very difficult. In the second place, the 
absence of inner logical conn -xion between the individual 
dicta of the Halakah made its study a work of mere 

of separating such traditional laws from the scriptural passages with which 
they had for centuries been connected. This would have remained their 
attitude even if they had realized that such a connexion was merely artificial 
(see below, note 27). No differences of opinion were therefore possible as to 
how such traditional laws were to be connected with the Scripture.- 

It should be noted that Hoffmann seems to have subsequently abandoned 
both his theories. In his introduction to his translation of the Mishnah, 
Seder Nezikin (p. x, note 3), he states that according to the Palestinian 
Talmud the so-called Number-Mishnahs were already compiled and redacted 
by the men of the Great Synagogue. He refers to the passage in Shekalim, 
V, 48c, which, like Weiss and Oppenheim, he misinterprets. See below, 
note 26. 

19 N. Krochmal {op. cit., pp. 174-5) a ' so assumes that even the last of 
the Soferim began to teach independent Halakot (so also Pineles, Darkah 
shel Torah, pp. 8-9). Like Frankel, Krochmal also gives as the reason the 
increased number of the Halakot and new decisions which could no longer 
be connected with the Scripture in the form of the Midrash. There is, 
however, a great difference of opinion between Krochmal and Frankel as to 
dates. Krochmal extends the period of the Soferim until about 200 B. c, 
assuming that the Simon mentioned in Abot as ' one of the last survivors 
of the Great Synagogue' is Simon II, the son of Onias II. Krochmal 
therefore designates him as the last of the Soferim and the first of the 
Mishnah teachers, the Tannaim {loc. cit., p. 166). According to Frankel, 
the last member of the Great Synagogue was Simon the Just I, about 300 b. c. 
This Simon, then, was the last of the Soferim in whose days the Mishnah 
was introduced {Hodegelica, pp. 68 and 30-31). 


mechanical memorizing — a very tiresome and repulsive 
procedure for the intelligent student. In the third place, 
the Pentateuch gives the laws pertaining to one subject in 
many different places. As the Midrash follows the Penta- 
teuchal order, there could be no systematic presentation of 
all the laws on any one subject. The laws on one subject, 
for instance, Sabbath, being derived from widely separated 
passages in the Pentateuch, had to be taught piecemeal, 
each decision in connexion with its scriptural basis. For 
all these reasons, Frankel tells us, the last group of the 
Soferim decided to separate the Halakot from their 
scriptural bases and to teach them in the new Mishnah- 
form systematically arranged according to subjects. 

Like Hoffmann, Frankel assumes that the plan of 
arranging the Halakot according to subject-matter was 
coincident with the very introduction of the Mishnah-form, 
so that the very earliest Mishnah collections must have 
been arranged topically. This, as we have seen, is in- 
correct. The topical arrangement of the Mishnah is of 
later date. It was preceded by other forms of grouping 
peculiar to the earlier Mishnah collections. Frankel him- 
self credits R. Akiba with the systematic arrangement of 
Halakah according to topics (pp. cit., p. 115). He also 
qualifies by the following remarks his former statement 
concerning the Soferim and their arrangement of the 
Halakah according to subjects: 'We have stated in the 
preceding chapter that the teaching [of the Halakah] 
according to subjects began at the end of the period of 
the Soferim. Nevertheless, a long time undoubtedly 
passed before all [the Halakot] that belonged to one 
subject were brought together under one heading. Very 
often while dealing with one subject they would [not keep 
VOL. V. Mm 


strictly to it but] drift to another and pass from one 
halakic theme to another . . . . R. Akiba, however, began 
to arrange the old Halakot to put each in its proper place 
and [under the topic] to which it belonged.' 20 If, however, 
the order in the Mishnah before R. Akiba was not strictly 
according to subjects, as Frankel here admits, and if some 
Halakot bearing on one subject would often be treated 
among Halakot dealing with another subject, what ad- 
vantage was there then in separating the Halakot from the 
Midrash and teaching them in the Mishnah-form ? The 
shortcomings of the Midrash-form, according to Frankel, 
consisted in the fact that the Halakot of one subject could 
not be taught connectedly but were interrupted by Halakot 
belonging to another subject. However, according to 
Frankel's own statement, the same defect was inherent 
in the Mishnah-form up to the time of Akiba. 

Taking up another statement of Frankel, it seems 
difficult to realize why the study of the written laws 
together with all the Halakot derived from them, as is 
done in the Midrash-form, should be such dry mechanical 
work of the memory, and so repulsive to the intelligent 
student. One would be inclined to think that the study 
of the Halakot in the abstract Mishnah-form, especially 
when not arranged systematically, would indeed be a far 
more mechanical work and far more tiresome for the 
student. Again, according to Frankel, it was the alleged 
lack of inner logical connexion between the single Halakot 

20 spD ^nnn awj/n a"y mcbn a*npn pnsa uaro -oa rani 
a^trn 5»a isdnj cna avn av/> i-djj pdd vta but ,anaiDn w 
latfDO m jyjn apDjn a^oya mini ,ihx bit nnn nnx pay ba 
by nwn rro^nn -nob bnn y"-n , . . nsbrb nsbna) pjj£ pjyo 


which made the Midrash-form inadequate for teaching 
purposes. However, this absence of inner logical connexion 
is merely alleged by Frankel, but not proved. If we should 
even grant that in the Midrash-form the Halakot were not 
always logically connected and coherently presented, the 
earlier Mishnah certainly did not remedy this evil. The 
earlier Mishnah collections were characterized by the most 
arbitrary modes of arrangement. Halakot bearing upon 
different themes and altogether unrelated in subject-matter 
were often grouped together under artificial formulas. 
Examples of these earlier modes of arrangement have been 
preserved even in the present form of our Mishnah as, 
for instance, in the so-called Number-Mishnahs or the 
En-ben-Mishnahs. The Midrash-form certainly established 
a better connexion between the individual Halakot than 
did these earlier arrangements of the Mishnah. The mere 
fact that many Halakot belong to one and the same- 
chapter or are grouped around one and the same passage 
of the Scriptures, establishes a better connexion between 
them than the accident that they can all be presented 
under one formula. 

Aside from all these arguments, the fundamental position 
of Frankel can hardly be maintained. In the time of the 
last group of the Soferim, the halakic material could not 
have grown to such an extent as to make it impossible 
to use the Midrash-form and necessitate the innovation of 
a new form of teaching. The mere volume of the halakic 
material could by no means have brought about this change 
of form. This is evident from the fact that our halakic 
Midrashim, Sifra, Sifre, and Mekilta, present in Midrash- 
form a mass of halakic material far greater in volume 
than was extant in the days of the Soferim. Thus we see 

m m a 


that all the reasons which Frankel gives for the introduction 
of the Mishnah-form are insufficient and could not have 
been the cause of the innovation. 

In conclusion, Frankel's admission that the teachers 
continued to use the Midrash-form even after the intro- 
duction of the Mishnah-form 21 is the strongest refutation 
of his own theory. If the Midrash-form had so many 
disadvantages, if it was both tiresome for the student and 
inadequate for presenting the Halakot systematically, why 
was it not altogether abandoned ? How did the new form 
obviate the evils of the old form if the latter continued 
in use? 

The theory propounded by Weiss in his Mabo la-Mekilta, 
pp. iv and v, and in his Dor, I, p. 66, is somewhat of an 
improvement upon the ideas of Frankel. Like Frankel, he 
believes that the Mishnah-form was introduced by the later 
Soferim, and that the reason for this change was the large 
increase of halakic material. He avoids two of the 
mistakes that Frankel made. In the first place, he does 
not confuse the innovation of teaching detached Halakot 
in the form of Mishnah with the arrangement of the 
latter according to subjects. Nor does he assume that 
the Midrash-form continued in use, after the Mishnah- 
form was introduced. According to Weiss, the Midrash- 
form was abandoned because it proved inadequate. It 
was hard for the student to remember the great mass 
of Halakot that existed at that time, when taught 
in the Midrash-form. The teachers, therefore, felt the 
need of inventing another form which would help the 

21 Op. cit., p. 7, he says : pJJ» DDXS& *|TI Dr6 VTV3B> f[« 'O SHI 

anpvh onows an tn wn hztt •nmb bwk-wi th my $b raf?n. 


memory retain the increased number of halakic teachings. 
This help for the memory they found in separating the 
Halakot from their scriptural bases and in expressing them 
in short, concise phraseology, and in arranging them 
according to a number-formula. The saying of Simon 
the Just, ' The world rests upon three things, &c.' (Abot 
I, 2), and the three Halakot mentioned in Eduyot VIII, 4, 
which according to Weiss are soferic Halakot, merely 
reported by Jose ben Joezer, are cited by Weiss in support 
of his theory that the Soferim taught detached Halakot 
expressed in concise terms and arranged according to 
number formulas. Weiss {Mabo la-Mekilta, p. v, note 7) 
admits, however, that the innovation was unsuccessful. 
The teachers, he tells us, soon found that the Mishnah- 
form, although superior to the Midrash, in being more 
easily memorized, had many other disadvantages. As 
a result, they had to return to the older form of the 
Midrash after they had abandoned it for a time. 22 

This admission of Weiss that the advantages expected 

22 In this assumption, that the Midrash-form had for a long time been 
abandoned and supplanted by the Mishnah, and that later on objections 
to the Mishnah-form caused a return to the Midrash, Weiss is followed 
by Oppenheim ('Ha-Zuggot we-ha-Eshkolot ' in Hashahar,V\l, pp. 114 and 
116), and by Bassfreund (see above, note 15). It is strange that while these 
scholars cannot account satisfactorily for one change that really took place, 
namely, from the exclusive use of the Midrash to the admission of the 
Mishnah-form, they assume another change which never took place, namely, 
a return from a supposed temporary exclusive use of the Mishnah to the 
old Midrash. We have already seen that the Mishnah-form was never in 
exclusive use, for the Midrash continued to be used side by side with it. 
Consequently there could have been no return from Mishnah to Midrash. 
But we shall see that the very reason which Weiss, Oppenheim, and 
Bassfreund give for the return to the Midrash, namely, the opposition 
of the Sadducees, was rather the cause for the further departure from the 
Midrash-form and the extension of the use of the Mishnah-form (see below, 
notes 72 and 73). 


from the new form were not realized, is in itself a strong 
argument against his theory. Further, we have seen above 
that the necessity for aiding the memory could not have 
been the reason for introducing the Mishnah-form. The 
words of the scriptural text with which the Halakot were 
connected in the Midrash-form offered sufficient help to 
the memory. We have also seen above that in the days 
of the Soferim the halakic material was not so large as to 
necessitate new forms and arrangements. The Soferim 
never gave their teachings in any other form but in the 
Midrash, namely, as interpretations and additions to the 
written laws. They never arranged them in any other way 
except in the order of the scriptural passages to which 
they belonged. The two passages, cited by Weiss, do 
not refute this statement. The saying of Simon the Just 
in Abot is not a halakic teaching but a maxim of the 
same character as the other wisdom literature of that time. 
We can draw no conclusions from it as to the form of 
halakic teachings of that day. As for the three Halakot 
mentioned in Eduyot, these will later be shown to have 
been the decisions of Jose ben Joezer himself. Conse- 
quently they do not prove anything concerning the form 
of halakic teaching used by the Soferim. 

Oppenheim 2S offers a theory that is in reality but 
a combination of the views examined above. However, 
he makes a very correct observation concerning the date 
of the innovation. According to Oppenheim, the Mishnah- 
form was first introduced during or immediately after the 
Maccabean uprising. As a result of the persecutions 
incident to the Maccabean revolution, the study of the 

23 'Toledot Ha-Mishnah' in Beth Talmud, II, p. 145, and also in his 
' Ha-Zuggot we-ha-Eshkolot ' in Hashahar, VII, pp. 114-15. 


law was neglected and the knowledge of it decreased. 
The teachers, therefore, decided to separate the Halakot 
from their scriptural bases and to teach them indepen- 
dently, in order to save them from oblivion (' Toledot ha- 
Mishnah', in Beth Talmud, II, p. 145). They chose this 
form either because they thought that in this form it 
would be easier for the student to remember the Halakot, 
or because they, the teachers themselves, no longer 
remembered the scriptural bases for many Halakot. 

The first of these two reasons is identical with the one 
given by Frankel and by Weiss, which has been found 
insufficient. The second one is similar to the one given 
in Hoffmann's second theory, and, as we have seen, is 
not plausible. For, if they had not previously studied 
Mishnah but received the Halakot only together with 
their scriptural bases, it is hardly possible that the 
teachers could forget the latter and yet remember the 
former. The remembered Halakot would have recalled 
to them the scriptural passages in connexion with which 
they were received. 

It seems that Oppenheim himself felt that neither his 
own nor Frankel's nor Weiss's theory was sufficient to solve 
the problem. He therefore offered another solution of the 
problem, and this is practically a denial of the fact that 
there is a problem. After stating that the Soferim taught 
in the Midrash-form and those who followed them intro- 
duced the new form of abstract Halakot, that is Mishnah, 
he contradicts himself by adding the following remark: 24 
' But in my opinion there is no doubt that the Soferim who 
taught [the Halakah] as a commentary on the Scriptures 

□T> llTOn fc6 m^nn \0 < Ha-Zuggot weha-Eshkolot', I.e., p. 114. 


[i. e. Midrash] also taught independent Halakot.' He then 
proceeds to prove that the Soferim had independent or 
abstract Halakot in the form of Mishnah. 26 According to 
this statement there is no problem at all. We need not 
account for any change in the form of teaching Halakah or 
explain the reasons for the innovation of the Mishnah, for 
there was no change and no innovation. The two forms, 
Midrash and Mishnah, were evidently used together from 
the earliest times, the Midrash possibly to a larger extent 
than the Mishnah. This would indeed be the best solution 
of the problem and would remove all difficulties. The only 
obstacle in the way of its adoption is that it is contradicted 
by all historic reports. It is against the tradition that in 
earlier times all the teachings of the Halakah were given 
in the Midrash-form only. This tradition, we have seen, 
is indicated in the discussion of Jose and Hezekiah men- 
tioned in the Palestinian Talmud (Moed katan) and is 
expressly mentioned by Sherira Gaon. It is also out of 
harmony with the generally accepted opinion that the 

26 This is also the stand taken by Halevi who goes even further and 
maintains (Doroth ha-Rishonim, I, chap, xiv, pp. 204 ff.) that in the main 
our Mishnah had already been composed and arranged by the Soferim, but 
he does not prove his statements. At the most, his arguments could only 
prove that there had been many Halakot and decisions in the days of the 
Soferim, and that the earliest Tannaim in our Mishnah in their discussions 
seek to define and explain these older Halakot and decisions. But it does 
not follow that these Halakot and decisions were already in the days of the 
Soferim composed in the Mishnah-form. These Halakot and decisions were 
originally given in the Midrash-form, as definitions or interpretations of 
written laws. The later teachers, that is, the earlier Tannaim, discussed 
and commented upon these decisions and Halakot of their predecessors 
which they had before them in Midrash-form. Later on, when these decisions 
and Halakot became separated from the Midrash, they were arranged in 
the Mishnah-collections as independent Halakot, together with all the 
comments and explanations given to them by the Tannaim, and in this form 
they are also found contained in our Mishnah. 


Soferim, as the name implies, imparted all their teachings 
only in connexion with the written book of the Law. It 
is, further, against an absolutely reliable report in the 
Babylonian Talmud which, as we shall see, tells us not 
only that the older form of teaching the Halakah was the 
Midrash, but also gives us the period of time during which 
it was in exclusive use. 

Thus we see that all these theories examined above have 
not succeeded in finding a real solution for our problem. 
None of the theories have given the exact time or the real 
cause for the introduction of the Mishnah-form. 

Probably the strangest feature of the problem is the 
silence of the talmudic literature about this important 
innovation. This silence is all the more remarkable when 
we come to realize that this was not merely a change in 
form, but an innovation that had great influence upon the 
development of the Halakah and had great bearing upon 
the validity of its authority. 

The theory proposed in this essay offers what appears 
to us to be a satisfactory solution for this many-sided 
problem. In the first place it determines the exact time 
when the innovation of teaching independent Halakot was 
introduced. In the second place it describes the conditions 
that compelled the teachers to make so radical a change. 
And finally it explains why no explicit report is preserved 
in talmudic sources regarding this great development in 
the teachings of the Halakah. This theory I shall now 

{To be continued.)