The Vocalization and the Reading Tradition
The next component of the Tiberian Masoretic tradition that we shall consider is the vocalization. This consists of a set of signs that were written below, above and sometimes within the letters of the consonantal text. The vocalization system includes signs to represent vowels and also signs to represent syllable division (shewa), consonant gemination (dagesh), the distinction between the two types of pronunciation of the so- called bgadkfat consonants (dagesh) and the consonantal pronunciation of a letter (mappiq). The vocalization notation, in fact, marks more than phonology. It reflects syntactic divisions in the same way as the accents, in that it marks differences between the pronunciation of words that occur at syntactic pauses and those that occur within syntactic units. The dagesh sign is sometimes used, moreover, in an exegetical function to distinguish meaning. A few isolated cases of this are found in the Tiberian tradition, such as the dagesh in the mem of ‘Abimelech’ in Genesis 26: 1 to indicate that this is a different Abimelech from the one mentioned in chapters 20–1 , and in the lamed of the word lo (meaning ‘not’) when collocated with the homophonous word lo (meaning ‘to him’), for example Proverbs 26: 17 . This usage of dagesh is more widespread in the Babylonian vocalization.
As is the case with the accent signs, the vocalization signs are a written notation that was developed by the Masoretes to record a reading tradition. It is not possible to establish exactly when the vocalization and accent signs were created. Neither the vocalization signs nor, as we have seen, the accent signs are mentioned in Talmudic literature or in other sources from the first half of the first millennium CE. Jerome (346–420 CE) expressly states that the Jews do not use signs to denote the vowels. In the earliest Masoretic codices datable to the ninth century, however, the notation of the vocalization and accents is fully developed, so the first stages of its development are likely to have taken place at least a century earlier.
In the time of the Tiberian Masoretes and also for a certain period after their activities ceased both the Tiberian sign system and the Tiberian reading tradition were regarded as authoritative. The form of sign system that became standardized represents a fixed stage in the development of the notation. Some extant manuscripts with non- standard Tiberian vocalization preserve more primitive stages of its development and others exhibit more developed stages. In the standard Tiberian system a vestige of a more primitive stage of development can be identified in the vocalization of the qere (what is read) of the tetragrammaton (yhwh) with shewa on the aleph of 'adonai. One can compare this to the continuing use of the early Hebrew script to write the tetragrammaton in Qumran manuscripts that are otherwise written in square script (see above).
The other vocalization systems (Babylonian and Palestinian) exhibit various degrees of assimilation to the Tiberian system in the extant manuscripts. The Hebrew grammarians in the tenth and eleventh centuries, and also other learned scholars, all followed the Tiberian vocalization and the Tiberian reading tradition, which it reflected, whether they were resident in Palestine, Iraq, North Africa, or Spain. The Tiberian vocalization system soon became the standard one and replaced all other systems in the transmission of the Bible. The transmission of the Tiberian reading tradition, on the other hand, soon came to an end. It is not completely clear why this happened. For one or two generations after the last Masoretes teachers of the Tiberian reading tradition could still be found in Palestine, but not, it seems, in all Jewish communities. The Spanish grammarian Ibn Janah (eleventh century) expressed regret that in Spain there were no traditional readers and teachers (ruwat wa-'ashab al-talqin) with a first hand knowledge of the Tiberian reading. The reading tradition may have become extinct through lack of trained teachers. Whereas the signs of the vocalization system could be copied by any scribe in any community, the oral transmission of the reading which depended on a small circle of teachers could not keep abreast of the large expansion of the transmission of the written Tiberian tradition in manuscripts throughout the Jewish world. As a result, the Tiberian vocalization signs came to be read according to the various local traditions of Hebrew pronunciation, most of them heavily influenced by the vernacular languages of the communities concerned. It is only recently, by studying previously neglected medieval sources, that we have been able to reconstruct the original Tiberian reading tradition. This does not correspond to the descriptions that are found in modern textbooks of biblical Hebrew, all of which present a form of pronunciation that was not that of the Tiberian Masoretes.
In a large number of places the reading tradition that is reflected by the vocalization does not correspond to the consonantal text. In the majority of cases the divergence relates to the pronunciation of single vowels in a single word. Sometimes there is a difference in the whole word, as in 2 Kings 20: 4 where ‘the city’ is written and ‘court’ is read, or the division of words. In a few isolated cases the discrepancy amounts to omissions or additions of words or phrases, as in Jeremiah 31: 38 where ‘behold days’ is written and ‘behold days are coming’ is read. The Masoretes indicated in their marginal notes the places where these discrepancies occurred. There are approximately 1,500 of these notes. Some elements of the consonantal text are regularly read in a way that does not correspond to what is written. These are not marked in the Masoretic notes. The most common word where this occurs is the tetragrammaton (yhwh), which is read either as 'adonai (my Lord) or as 'elohim (God). It also applies to the reading of some elements of morphology. The regular discrepancy between the written form of ‘Jerusalem’ (yrwshlm) and the reading tradition with final -ayim is likewise a morphological difference.
There is no uniform trend in the deviations of the reading tradition from the consonantal text. In a few isolated cases the reading tradition replaces possibly offensive words with a euphemism; for example, 1 Samuel 5: 9 ketiv: ‘haemorrhoids’ (the meaning is not completely certain), qere: ‘tumours’. The avoidance of pronouncing the tetragrammaton, moreover, is presumably theologically motivated. In the vast majority of cases, however, the qere does not appear to be an intentional change of the written text.
The most satisfactory explanation for this phenomenon is that the reading was a separate layer of tradition that was closely related to, but nevertheless independent from, the tradition of the consonantal text. Contrary to a view that is still widely held today, the reading tradition was not a medieval creation of the Masoretes but was an ancient tradition that the Masoretes recorded by their notation system. This tradition had been faithfully passed on orally from teacher to pupil over many generations. There is no evidence that the Masoretes reformed the reading tradition and felt free to introduce exegetical or linguistic innovations of their own.
In the discussion of the history of the reading tradition we should distinguish its textual form from its linguistic form. There is evidence that both of these aspects have ancient roots. The textual differences between the reading and the written text are referred to in Talmudic literature. Some of the Qumran scrolls from the Second Temple period have in a number of places the text of the Tiberian qere. One may trace back the text of qere forms even further, into the period of literary growth of the biblical books. This is shown by the fact that the ketiv of the text of Chronicles often corresponds to the qere of its earlier biblical source. An example of this is the word migrasheha, ‘its surrounding pasturelands’, which is used in association with the lists of Levitical cities in Joshua 21 and 1 Chronicles 6 . The Chronicler is clearly using the text of Joshua 21 as his literary source. In the original text in Joshua the word is always written as a singular form but it is read in the reading tradition as a plural. This reflects a later interpretation of an originally singular form as a plural. This ‘later’ interpretation, however, is no later than the consonantal text of Chronicles, where it is written as a plural. Even if we do not attribute this interpretation to the author of the Chronicles passage, there are good grounds for arguing that the text of the reading tradition of Joshua 21 is as old as the consonantal text of 1 Chronicles 6 . Linguistic features of the Tiberian reading tradition that differ from what is represented in the consonantal text are reflected by some Qumran manuscripts. This is seen, for example, in the form of some of the pronominal suffixes.
As we have seen, in the Middle Ages various ways of pronouncing biblical Hebrew are reflected in different systems of vocalization. The Tiberian, Babylonian, and Palestinian systems of vocalization not only use different sets of signs but also reflect clearly distinct forms of pronunciation. Indeed in manuscripts within the Babylonian and Palestinian systems one can identify several varieties of pronunciation. In addition to these three traditions of pronunciation, there is the Samaritan tradition, which was not recorded in written notation but has been passed down orally. Although the Tiberian, Babylonian, and Palestinian systems differ from one another, it is clear that they are closely related in comparison with the Samaritan pronunciation of Hebrew, which is significantly different from all three. We can identify two broad streams of pronunciation tradition, the Samaritan and the non-Samaritan. The close relationship of the Babylonian reading tradition with the Tiberian and Palestinian could be explained as a result of its being transferred from Palestine to Babylonia by Jewish scholars after the Bar-Kochba revolt. These Palestinian scholars also established the first rabbinic academies in Babylonia at this time. Similarly the official Targumim of Onqelos and Jonathan appear to have been transferred from Palestine to Babylonia in the same period.
A number of the differences within the non-Samaritan group appear to have arisen through the influence of the vernacular languages. This applies especially to the Palestinian pronunciation, which exhibits many features that are characteristic of Aramaic, the vernacular of the Jews for most of the first millennium CE. A number of Aramaic features can also be identified in the Babylonian pronunciation of Hebrew, though it appears that it differed from contemporary vernacular Aramaic in a number of ways and was a conservative tradition. The Tiberian system appears to have been very conservative and was relatively unaffected by vernacular influence. The greater concern for conservatism in the Tiberian and Babylonian traditions is reflected by the fact that a corpus of detailed Masoretic annotations was developed by the Masoretes of Tiberias and Babylonia, but manuscripts with Palestinian vocalization exhibit only sporadic marginal notes, mainly concerning qere and ketiv. We may compare this to the varying degrees of conservativeness in the transmission of the Aramaic Targumim. Targum Onqelos of the Pentateuch preserves a form of literary Aramaic that was used in Palestine at the beginning of the first millennium CE. The text of this Targum was stable in the Middle Ages and Masoretic notes were developed to ensure its accurate transmission. The so-called Palestinian Targumim, on the other hand, reworked earlier Targumic traditions in the vernacular Aramaic of Palestine. Their text was by no means fixed and so no Masoretic notes were developed in association with their transmission. Another feature that reflects the concern for accurate transmission is the fact that the Tiberian Masoretic tradition developed a full system of vocalization, in which every word and virtually every letter had its vocalization sign, even if this denoted zero (shewa). Manuscripts with Babylonian and Palestinian vowel signs do not exhibit such a consistently full system. This especially applies to Palestinian vocalization, which is generally marked only sporadically on isolated words.
It was no doubt for this reason that in the Middle Ages the Tiberian reading tradition was the preserve of a small number of scholars who had received special training. The Palestinian pronunciation, which was close to the Aramaic vernacular, was far more widespread. The Sephardi pronunciation traditions of Hebrew, which are still followed today in many of the eastern Jewish communities, are derived historically from Palestinian pronunciation. The Babylonian pronunciation, which was also more widespread in the medieval Jewish communities than Tiberian pronunciation, has survived down to the present day in the reading traditions of the Yemenite Jews.
We have already discussed the evidence for the existence in the Second Temple period of certain textual and linguistic elements of the Tiberian reading tradition that differ from the consonantal text. The linguistic features in the Qumran manuscripts that correspond to the Tiberian reading tradition indicate that these features were not introduced into the reading tradition in the Middle Ages. There is also evidence that the Tiberian reading tradition resisted the influence of the Aramaic vernacular during its transmission in the first millennium CE. This is seen clearly in the reading tradition of the Aramaic portions of the Bible. In numerous places the reading tradition of biblical Aramaic reflects a different morphology from that of the consonantal text. This reflects the independence of the two traditions. The Aramaic morphology of the reading tradition, however, is not the same as the morphology of Jewish Palestinian Aramaic, the dialect that was spoken by Jews in Palestine throughout the Byzantine and early Arab period, but has earlier roots. Jewish Palestinian Aramaic was spoken by the Masoretes during most of the Masoretic period so this is evidence that the Tiberian reading tradition was not influenced by the vernacular speech of its transmitters. Indeed there are features of non-Semitic pronunciation in loanwords from non-Semitic languages that were preserved from the original period of composition right down to the period of the Masoretes, centuries after contact of the transmitters of the tradition with the source language had ceased. This demonstrates the incredible conservatism of the Tiberian reading tradition.
We have seen that some linguistic features of the Tiberian reading tradition are attested in the Qumran manuscripts. However, Qumran sources also reflect various features of phonology and morphology that are alien to the Tiberian tradition. This applies also to the reading tradition of Hebrew reflected in transcriptions of Hebrew words (mainly proper nouns) that are found in the Septuagint. Some of the distinctive linguistic features of the Samaritan tradition can be traced back to the Second Temple period.
During the Second Temple period, therefore, there were a variety of reading traditions of the Hebrew Bible which differed from one another both linguistically and also textually. The lack of correspondence of some forms of pronunciation with the Tiberian reading tradition should not lead us to conclude that the Tiberian tradition is a later development. There is evidence of the extreme conservatism of the Tiberian tradition and it is likely that a form of pronunciation that is very close to the Tiberian tradition existed in Second Temple times side by side with other traditions of pronunciation. The fact that transcriptions in the Septuagint, for example, often have an a vowel in an unstressed closed syllable (e.g. Mariam) where in Tiberian Hebrew it has developed into an i (Miryam) should not be interpreted as demonstrating the chronological antecedence of the Septuagint reading tradition, although it may reflect typologically an earlier stage of development. It is relevant to take into account that in the development of the dialects of a language some dialects may be more conservative of earlier linguistic features than other dialects spoken at the same period. Some features in the transcriptions of the Septuagint and other early sources that differ from Tiberian phonology can, in fact, be explained as the result of influence from the Aramaic vernacular, which was resisted by the standard Tiberian tradition. Likewise, where the Qumran biblical scrolls reflect a different pronunciation from the Tiberian one, it should not be assumed that the Tiberian is a later development. Some Qumran scrolls that are written according to the Qumran scribal practice, for instance, exhibit a weakening of the guttural consonants, whereas these are stable in the Tiberian tradition. It is clear that the Qumran scribes were influenced by vernacular pronunciation whereas the Tiberian tradition is conservative and has preserved the original distinction between the guttural letters.
Similarly, where the reading tradition of the consonantal text reflected by the Septuagint differs textually from the Tiberian, it does not necessarily follow that the Septuagint tradition is the original one and the Tiberian is a later development. There is, in fact, considerable textual agreement between the vocalization reflected by the Septuagint and the Tiberian one. This shows that there must have been a large degree of continuity in the reading tradition. The places where the vocalization adopted by the Septuagint translator differs from the Tiberian tradition can in some cases be shown to be the result of uncertainty and conjecture and so the Tiberian vocalization, although later, would preserve the older, more accurate tradition.
The precise relationship of the Tiberian tradition with the Babylonian and Palestinian traditions is not completely clear. There was a complicated web of relations between the traditions of reading the Hebrew Bible in the Second Temple period, just as there was between the various forms of the consonantal text of the Bible. As remarked above, the Babylonian and Palestinian reading traditions are more closely related linguistically to the Tiberian than to the Samaritan, yet some linguistic features of the reading tradition reflected by the Septuagint transcriptions are found in the Babylonian traditions but not in the Tiberian.
As remarked above, the Tiberian vocalization marks syntactic divisions by distinctive pausal forms of words. In the majority of cases the occurrence of these pausal forms correspond to the divisions expressed by the accents. In a few cases, however, they conflict with the accents, for example, in Deuteronomy 6: 7 (‘when you sit in your house and when you go on the road’). Here a pausal form occurs on ‘in your house’ but according to the accents the major pause should occur on ‘on the road’ not on ‘in your house’. Note also Deuteronomy 5: 14 ‘you, your son, your daughter, and your maidservant’, where a pausal form occurs even with a conjunctive accent (‘and your daughter’). Cases such as these suggest that the tradition of vocalization and the tradition of accents were independent from each other to some extent. We see, therefore, that the vocalization tradition is a layer of tradition that is not only separate from the consonantal text but also from the accents. So the reading tradition, which includes both vocalization and accents, comprises two separate layers.
The separateness of the reading tradition from the tradition of the consonantal text is reflected in the fact that in the Talmudic period different exegesis was applied to each layer. Some of the rabbis followed the principle that both the consonantal text and the reading tradition were authoritative sources. Rabbinic literature contains numerous examples of a different interpretation being made of the two levels of the text. This is also reflected in the rabbinic form of exegesis that is expressed by the formula 'al tiqre … 'ella ‘do not read … but.…’ The purpose is probably to express an interpretation of the consonantal text that differs from the reading tradition rather than to record a variant reading tradition. It is not clear how far back this exegetical practice can be traced.
Another sign of the independence of the reading tradition from the written transmission is the existence of transcriptions of the reading tradition into other alphabets. We have seen that in the Middle Ages the Karaite Jews transcribed the reading of the Hebrew Bible into Arabic script. They also occasionally transcribed the Hebrew of the Mishnah and other rabbinic literature, which themselves had an orally transmitted reading tradition, but never medieval Hebrew texts such as legal texts or Bible commentaries that had no reading tradition and circulated only in written form. At an earlier period the reading tradition of the Hebrew Bible was represented in Greek transcription. The clearest evidence for this is the second column of Origen's Hexapla, which contained a full Greek transcription of the reading of the Hebrew text. From internal evidence it appears that this transcription was taken from an earlier source datable to no later than the first century CE. Similar transcriptions of the reading tradition may have been used by Greek-speaking Jews in the Second Temple period, although there is no direct evidence to substantiate this hypothesis.
The recitation of the Hebrew Bible in Jewish worship is presented as an established custom in the New Testament (Luke 4: 16ff., Acts 15: 21 ). It is mentioned by Philo (c.20 BCE–50 CE) and Josephus (first century CE) and is likely to go back several centuries earlier. Public reading of the Pentateuch (or parts of it) is referred to in the Bible (Deut. 31: 9–11, Neh. 8: 1–8 ). It can be argued that the very existence of Bible manuscripts implies the contemporary practice of public recitation, since in the Second Temple period a large proportion of the Jewish population must have been illiterate. We have seen that the consonantal text of the Tiberian Masoretic tradition can be traced back to the earliest attested manuscripts from the Second Temple period (third century BCE) and, on the basis of orthography, can be carried back further, possibly to the time of Ezra. Rabbinic traditions concerning the recitation with accents are linked to Ezra. In the introduction to the medieval treatise Hidāyat al-Qāri (‘The guide for the reader’), the transmission of the Tiberian reading tradition is traced back to Ezra. These statements should, of course, be approached with caution and they cannot be verified. It is clear, however, that the Tiberian reading tradition has ancient roots. When the proto-Masoretic consonantal text was fixed it was already corrupted by scribal errors. It reflected, moreover, literary recensions of some of the biblical books that were different from what is found in other transmissions. In most cases the Tiberian reading tradition has been adapted to the words in the consonantal text that have been corrupted by earlier scribes. The reading tradition, furthermore, although deviating from the consonantal text in some places, does not reflect a radically different literary recension. On these grounds the reading tradition is unlikely to be older than the period in which the proto-Masoretic consonantal text was fixed. A number of linguistic features suggest that the vocalization should not be dated to an earlier period. Some archaic forms of Hebrew morphology that are preserved in the consonantal text, for instance, are harmonized in the vocalization to the standard form, which is reflected by the majority of the consonantal text. Furthermore, judging by north-west Semitic epigraphy from the beginning of the first millennium BCE, final case vowels were still in use in Canaanite languages in the period when the earliest biblical passages were composed, but these do not appear in the vocalization. It should be taken into account, however, that some of the later biblical books were probably composed after the process of fixing the proto-Masoretic text had started. Linguistic differences between these compositions and the earlier books are clearly reflected by the consonantal text. There is also some evidence that the historical layering is reflected by differences in the vocalization, and that in the main the linguistic roots of the vocalization of the Bible are in the Second Temple period.
Exegetical alterations can be found in the reading tradition, just as they can in the proto-Masoretic consonantal text. Examples of this practice are cases where an original expression of ‘seeing the face of God’ is changed into the theologically more acceptable ‘appearing before God’ by reading the verb as a niph‘al (passive) rather than as a qal (active), for example Deuteronomy 16 : ‘Three times a year all your males shall appear before the Lord, your God.’ This change is clear where the verb is an infinitive and it lacks the expected initial he of the niph‘al form in the consonantal text, for example, ‘When you go up to appear before the Lord, your God.’ This change in the reading tradition is reflected already in the translation of the Septuagint and the Targumim, which demonstrates that it has ancient roots.
The proto-Masoretic manuscripts from Qumran exhibit a basically homogeneous text, but are by no means identical in all details. The text of the reading tradition that became accepted as the standard can be regarded as an oral form of a proto-Masoretic text which differed in some details from the written form that became standardized. The linguistic form (phonology and morphology) of the reading tradition is likely to represent one of various types that existed in the Second Temple period.