VOWELS AND DOTS CLASSICAL HEBREW. Alan Smith VOWELS IN CLASSICAL HEBREW THE DOT IN THE HEBREW LETTER. Two corrective studies.

Transcription

1 VOWELS AND DOTS in CLASSICAL HEBREW VOWELS IN CLASSICAL HEBREW THE DOT IN THE HEBREW LETTER Two corrective studies Alan Smith Elibooks

2 Part 1 First edition (draft) printed Shevat 5761 Standard edition printed Iyyar 5767 Five lines added to the bottom of page 8 Iyyar 5771 Part2 Alan Smith Revised Iyyar 5711 (2011) with mainly minor changes to the wording and the addition of many more examples

3 CONTENTS Part 1 Vowels in Classical Hebrew Prefatory Note The Prehistory of Hebrew Vowels Silent Letters The Masoretes The Tiberian Vowels Other Markings The Sheva and Hataf Syllables The Sephardi Vowels The Kimchi System Greek & Latin Vowels Long Vowels & Short Vowels (3 concepts) The Babylonian System Return to Tiberian Lengthening & Shortening of Vowels The Effect of the Missing Dagesh Mediaeval Hebrew Poetry Primitive Scribal Customs Masoretic Conventions Scribal Scribble Requiring Incineration Diphthongs Pausal Forms The Hataf-Kamets Precise Pronunciation of Vowels Rules and Guidelines Conclusion Appendix A Vowel Inter-connection Charts Appendix B The Babylonian System

4 Part 2 The Dot in the Hebrew Letter The shuruk Other dots The heritage of scribal laziness The strong dagesh The weak dagesh The mappik (a) The mappik in hé (b) The mappik in alef (c) The mappik in yod, and others (d) The dehik (and atey merehik) Exceptions with the letter> Unique exceptions Appendix The Dot Over the Letter The andµ The Holem The Mysterious Dot

5 Part 1 VOWELS IN CLASSICAL HEBREW Prefatory Note This is not for the average layman. Nor is it essentially for the scholar. Many people who are not scholars or academicians are interested in classical Hebrew grammar, and have tried to reconcile what they have been taught about the vowels with what they find in the text. Their lack of success is generally because what they have been taught does not agree with the text. What we were all taught is based on what is called the Kimchi system, devised and developed by the famous Kimchi family of grammarians, Rabbi Joseph and his sons Rabbi David (known as Radak) and Rabbi Moses, some seven to eight hundred years ago. The system presumes that originally Hebrew had, apart from the sheva, ten vowels five short and five long. The Ashkenazi pronunciation has preserved this, apart from pronouncing the long ā as a short o, while the Sephardi pronunciation has simply shortened all the vowels. This theory was intended to explain and simplify many rules of grammar. Unfortunately there is no historical basis for it, no real evidence to support it, and worst of all, it does not work. To a small extent it simplifies some parts of Hebrew grammar, but in a complex way, and has led to grammarians devising rules based on it which also do not work. Rather than admit that the rules are wrong, sub-rules and sub-sub rules are invented and all sorts of other concepts devised in order to squeeze what is found in the text into the rules the grammarians have devised. Some centuries ago, physicists formulated the Law of Conservation of Energy, that Energy cannot be created or destroyed. Then, time and again, they found that it didn t work, and each time they invented a new form of energy (or described some additional phenomenon as a form of energy) so as to make it work! (Einstein even went so far as to say that all matter is merely a form of energy.) In one case where they could not do this, they invented something completely fictitious called potential energy (i.e. there isn t any energy, it has disappeared completely, but it is capable of re-appearing), and all-in-all so far they have managed to get away with it. That is science. Grammarians tried the same thing with the Hebrew language. Based on the Kimchi system they invented a number of grammatical laws, then each time they found that one of them didn t work they either ignored the fact, or, if they could not get away with that, they invented a new grammatical concept, simply to make

6 6 Vowels in Classical Hebrew their ridiculous rules work! The worst case was the invention of a type of sheva that is something in between a silent (nach) and sounded (na) sheva. This concept, never previously considered, is actually taught in Israeli schools by order of the Ministry of Education! Not only does this wreak havoc with simple traditional grammatical concepts that do work (or did until this nonsense was thought up), but the results are so complicated that they hinder the understanding of the workings of the language, instead of helping it as grammar is intended to do. The amended Law of Conservation of Energy, if you allow potential energy, does (so far) work. The crazy amended grammatical rules based on the Kimchi system do not. We need to go back to the beginning or at least as far back as we can to examine properly what happened. We may find in the end that we have to replace simple rules that do not work, or that work only in a complex way by squeezing and distorting the text and adding lots of sub-rules, with new rules that are not quite so simple, but which in so far as they go, do actually work. The Prehistory of Hebrew Vowels. Hebrew writing from the time of Moses did not include vowels. This does not mean that there weren t any, or that they didn t pronounce them, but merely that they did not bother to write them. This was not an Israelite eccentricity. It is the way things were done by all other nations who used the Hebrew language at the time, as well as those who used many other languages. The Hebrew alphabet was very similar to that of the Phoenicians, from whom the Greeks took their alphabet, but the Greeks added symbols for vowels, and the Romans who followed them did likewise in Latin. Through a long tradition we know roughly how the Greeks and Romans pronounced their vowels. The earliest clues that we have as to how Hebrew was pronounced some two thousand years ago comes from the Greek translations of the Bible about that time. Translation does not help in itself, but names of people and places were not translated, they were transliterated, and from this we can know the Greek way of pronouncing them, and assume that the Hebrew was similar. By that time the Jews were already dispersed, the main centres being Israel, Babylonia and Egypt. Long before they were dispersed, in the time of Jephthah, we know that in the Land of Canaan itself different dialects existed, to such an extent that even consonants were pronounced differently (shiboleth sibboleth) in different parts of the country!

7 Vowels in Classical Hebrew 7 Even today, after some seventy years of radio and two wars with national military conscription to facilitate standardisation in England, many words are still pronounced quite differently in the North of England and in the South. The distance between London and Manchester or Leeds is far less than that between Jerusalem and Alexandria or Babylon. Furthermore, the native tongues of those places (Aramaic and Egyptian Greek) were quite different, and the native tongue always has a big influence on how the local Jews pronounce Hebrew. The Greek transliteration was probably based on the way Hebrew names were pronounced in Alexandria, Egypt, at the time. There is nothing to suggest that the Jews in Babylon used the same vowels. Until the coming of the Masoretes, there was no notation for writing vowels in Hebrew. The Masoretes, about whom we have a lot more to say, lived between a thousand and fifteen hundred years ago, and developed notation for representing vowels. It was only after that that the Hebrew language began to be studied seriously as a language with particular reference to its grammar. Silent Letters One point however needs to be mentioned before we proceed. The early grammarians and grammatical scholars (after the Masoretes, but earlier than the Kimchis) understood many of the principles of grammar extremely well, but were handicapped by having no suitable terminology with which to explain themselves. They did their best. Those who followed them developed their ideas, but also failed to understand some of them, which they distorted. The distorted ideas persist until today. One point in particular is that of the silent letters. The letters -)($are often silent, and tend to indicate which vowel precedes them, or at least to limit the choice of the possible vowels that precede them. They were often used, though not very much in traditional sacred texts, to indicate the presence of such vowels. But they are not themselves vowels. The later grammarians teach us that they are often used as vowels, which is just not true. They merely indicate, sometimes, the nature of the preceding vowel. The early grammarians describe them, when used this way and not as normal consonants, as ¼- I+L4 which means resting, i.e. silent. However, the Kimchis in particular regarded these silent letters as an integral part of the vowel, a concept we must discard as soon as possible. We mention this here, because this use of the four letters mentioned above to indicate (but not to represent) vowels is pre-masoretic. If you cannot quite grasp this, do not worry, we will explain it better later on.

8 8 Vowels in Classical Hebrew The Masoretes The Masoretes were a group of scholars living in the period between a thousand and fifteen hundred years ago, who set about preserving not only the text of the written classical Hebrew language but also the way it was to be read, at a time when it was no longer spoken. Here we are not concerned with listing all the valuable work they did, merely that in the course of time they produced notation for adding vowels to existing text without adding any extra letters (which would alter the text). Before explaining how they set about this, let us take a look at the situation at the time. There were three main centres of Jewish population, and of Jewish scholarship, and therefore disregarding all minor dialect variations there were three main recognised dialects of Hebrew. The first centre was in the Land of Israel, where the scholars who maintained the traditions of pronouncing Hebrew carefully were concentrated in the City of Tiberias. Their way of speaking Hebrew was what we refer to as the Tiberian pronunciation. The second was in Babylon, and the third in Egypt. Evidence seems to imply that the Tiberian pronunciation developed in the course of time into what we know as the Ashkenazi, and the Egyptian into the Sephardi. For historical reasons outside the scope of this book, the Jews of Babylon (later Irak) eventually lost nearly all of their old traditions, including the way they pronounced Hebrew, and adopted the Sephardi systems instead; but the old Babylonian traditions were preserved until very recently (and some to this very day) by the Yemenites. The systems (note the plural) of vowel notation developed by the Masoretes were not suddenly laid down. Someone suggested a system and it was adopted? No, various ideas were used, at first only some vowels were marked, then others, and in different ways. Some systems just did not survive we have odd manuscripts that use them, but they were not adopted generally, and died out. Gradually two systems evolved, were developed and standardised, and then adopted. One of these systems was developed by the group of Masoretes living in Tiberias, it represented the accepted Tiberian pronunciation, and is the one we all use to this very day. It also matches, fairly well, the Ashkenazi pronunciation (but not the Polish). The vowels are represented by dots, dashes and other marks placed normally underneath the letters, except for one that sometimes appears in the middle of a letter vav and one that appears on top.

9 Vowels in Classical Hebrew 9 The other system was developed by the group of Masoretes living in Babylonia, it represented the (then) accepted Babylonian pronunciation, and was in general use in the Yemen until quite recently. Many Yemenites can still read it. It was driven out only when they started using printed books instead of relying entirely on manuscripts; there was no printing in Yemen, books were imported and were only available in the Tiberian system, so the Yemenites had to learn that and adapt to it. The system matches the Yemenite pronunciation, which is not so very far removed from the Ashkenazi as far as vowels are concerned. The vowels are represented by symbols placed normally over the letters. We are fortunate in that we have copies of the identical text (the Bible) available in both systems. Thus we discover that there is a correspondence of vowel symbols between the two systems, except for the segol which the Babylonians did not have. (Nor do the Yemenites use it in pronunciation to this day.) The Egyptian and subsequently Sephardic pronunciation of the vowels was quite a lot different, but there appears to have been no group of Masoretes there to record it. Unfortunately, therefore, there is no notation to represent the Sephardi pronunciation. There is evidence that someone once tried to invent one, but it did not catch on. The result is that the Sephardim, when they found the need to write down vowels, used the Tiberian system, which is totally unsuited to their pronunciation. The system distinguishes between vowels which are the same to the Sephardim (e.g. segol and tserey) but does not distinguish between different Sephardi vowels which were the same to the Tiberians, specifically the kamets. Hence the problem of the short and long kamets. To the Tiberians (and modern Ashkenazim) a kamets is a kamets, pronounced roughly o, and that is all there is to it, so one symbol suffices, but to Sephardim there is an important difference between the two. Although we are concerning ourselves essentially with the Tiberian vowels, we will need to refer occasionally to the Babylonian and the Sephardic systems in order to help us understand what is going on, and to understand and correct wrong concepts, especially those developed with the best of intentions by the Kimchis. To understand how the latter arose, we may also need to consider the vowels used in Greek and Latin, bearing in mind that the latter at least were certainly familiar to the Kimchis.

10 10 Vowels in Classical Hebrew The Tiberian Vowels The earliest grammars refer to the seven vowels in the Tiberian system. (Modern grammars that teach ten are based on Kimchi the old ones teach seven.) Here are the vowels and the way they are represented. 1. i Represented by a single dot under the letter. Hirik H$. 2. é Represented by a horizontal pair of dots under the letter. Tserey I$. 3. è Represented by a triangular group of dots under the letter. Segol J$. 4. a Represented by a horizontal line under the letter. Patach K$. 5. o Represented by T under the letter. Kamets L$. 6. ó Represented by a dot over the letter (later moved to the left) Holom u$ N$. If followed by a silent vav, the dot is placed over the vav instead. }$. 7. u Represented by a diagonal row of three dots under the letter. Kubbuts O$. If followed by a silent vav, a dot is put inside the vav instead. Shuruk $. Note that the kubbuts and shuruk are the same vowel, only written differently. If a silent vav follows, the shuruk is used; if not, the kubbuts. There is no difference in pronunciation. There are, in addition, the sheva and the three hataf vowels (half-way between a sheva and a full vowel)g$ F$ E$ which we will deal with shortly. Other Markings The Masoretes introduced other marks under, inside and above the letters. 1. The te amim for punctuation. 2. The dagesh kal which indicates a choice of pronunciation of certain letters. 3. The right dot and centre (later changed to left) dot to distinguish shin and sin. 4. The mappik dot to indicate that a normally silent letter, especially the letter hé at the end of a word, is not silent. 5. The rafeh sign, the opposite to a dot in a letter (later dropped by the scribes). 6. The dehik (and atey merehik) to ensure that words are separated in pronunciation. 7. The meteg (or gaya) a complex subject on its own. 8. The strong dagesh, a dot in the middle of a letter to indicate that the letter is doubled.

11 Vowels in Classical Hebrew 11 The only one we are concerned with here is the last. It had always been the tradition, when a letter is doubled, to write it only once, even though it is pronounced as double (like the Italian bella). In such a case, they placed inside the letter a dot known as the dagesh hazak or strong dagesh. Here we will not indicate how this is distinguished from other dots, just that it exists. The Sheva and Hataf There are two types of sheva (represented by a vertical pair of dots under the letter. D$). The first is the silent sheva, which is really a non-vowel. It is placed under any consonant that ends a syllable, i.e. is not followed immediately by a vowel, but is not placed under a silent letter (silent ). The scribes decided to omit it also under the last letter of a word, except if it is a., or a º (with dagesh) or there are two consonants together. The second is the sounded sheva, pronounced like the phonetic. (Examples in English are ago, better, potato pronounced go, bett, p tato.) This follows a consonant that opens a syllable and is followed by another consonant and then a vowel in the same syllable. A sounded sheva cannot occur under the letters+(6$. It is replaced by a hataf G$ F$ E$. If the sounded sheva is a half-vowel, the hataf is a three-quarter vowel, or if the sheva is a quarter-vowel the hataf is a half-vowel anyway it is in theory very short, and behaves in grammar exactly like a sounded sheva. The hataf kamets G$ sometimes has another use, which is considered later. The hataf patach F$ is sometimes used in special cases in some MSS and editions to merely emphasise that a sheva is sounded (against instinct), e.g. ¼ªF/ J<L% D- B¼ ¼2F2¼} ¼< and in Gen. 21:6 only, ¼; K+F:H-. Long after the Masoretes, someone invented a floating sheva (sounded yet silent!) to justify a rule that does not work. Leading grammarians rejected it, but sadly the Gaon of Vilna, and, alas, the Israeli Ministry of Education, were persuaded to accept it. (Considered later under requiring incineration.) Syllables (A brief summary) A Hebrew syllable always begins with a consonant. (Forget the initial that replaces D) that is a special case that breaks all the rules and must be dealt with on its own.)

12 12 Vowels in Classical Hebrew A simple syllable starts with a consonant (C), followed by a vowel (V). If this is followed by a consonant which has a silent sheva underneath it (or should have see later), the second consonant closes the syllable, and we have a closed simple syllable (CVC). Otherwise, the syllable is considered as open, ending with either a vowel or a silent consonant (CV). Note that a silent consonant (which can occur in the middle or at the end of a syllable but can never open a syllable) does not count at all, and when considering syllables it is treated as if it were not there. A strong dagesh doubles the letter, and therefore closes the syllable and opens the next one. For example, ¼'I H0 is treated as if it were ¼'I2D2H0, ¼>LxĶ is treated as if it were ¼>LxD%Ķ and so on. A compound syllable starts with a consonant (C) followed by a half-vowel (v), viz. a sounded sheva or hataf, followed by a second consonant followed by a vowel. If this is followed by a consonant with a silent sheva, we have a closed compound syllable (CvCVC), otherwise an open compound syllable (CvCV). At the end of a word, a syllable may sometimes be closed by two consonants, but otherwise two consonants can never occur together in the same syllable, without a vowel or hataf or at least a sounded sheva between them. These are the only types of syllable possible in Hebrew. The only exceptions to the above are the words ¼1L6U'D;L- ¼0I$D>D;L- ¼-IºḐ ¼1¼-IºḐ ¼1H-KºḐ which must be treated as special exceptional cases. The Sephardi Vowels The Sephardi system has only five vowels: i e a o u, all short. The sheva and hataf are as in the Tiberian system, except that the hataf is pronounced as a full vowel. i is H$ è is J$ or E$ or I$ (no difference) a is K$ or F$ or (most times) L$ o isn$ or}$ or G$ or sometimes L$ u is O$ or $. Remember that the signs were not designed with the Sephardi pronunciation in mind, and they are most unsuited to it.

13 Vowels in Classical Hebrew 13 The Kimchi System Although this does not work, it must be explained because the reader has almost certainly been taught this, not under that name but as basic Hebrew grammar. In trying to reconcile how the two different systems (Ashkenazi and Sephardi) of pronunciation came about, Rabbi Joseph Kimchi suggested that there were originally ten vowels, five long and five short: i e ē a ā o ō u ū He also takes into account full spelling (Hebrew ¼$I0 L2) where a vowel is followed by a silent letter (vav or yod) and brief spelling where the same vowel is not. H$ ( brief, i.e. not followed by a silent yod) is i, unless it is an exception ¼-H$ ( full, note the silent yod) is, unless it is an exception J$ is e I$ (or¼-i$) is ē K$ is a L$ is usually ā, sometimes o N$ (or}$) is ō O$ ( brief, i.e. not followed by a silent vav) is u, unless it is an exception $ ( full, note the vav) is ū, unless it is an exception. Then, say the Kimchis, all that happened is that the Sephardim shortened all the vowels, while the Ashkenazim did not, but simply changed ā to o, and hey presto! It looks beautifully simple. It also explains how in certain circumstances a short vowel lengthens to the corresponding long one, and a long vowel in other circumstances shortens to the corresponding short one. Unfortunately there are far, far too many exceptions. Each of these has to be explained in some way, and if it cannot be explained with a midrash then we assume that the author had some secret reason for doing it that way, against the rules of grammar. Also on this are based certain rules about long vowels and short vowels in open and closed syllables having to take or not having to take the accent. But these, like many rules devised by the later Hebrew grammarians, are based on wishful thinking. You think up a rule that is nice and simple, and then squeeze everything in to fit in with it. You dare not (heaven forbid!) adjust the rules of grammar to fit in with what we actually find! The Tiberian system is based on seven vowels, and we find this number referred to in the earliest grammar books (such as Dikdukey Hate amim), not ten. There is

14 14 Vowels in Classical Hebrew no mention of long and short vowels, and there is evidence, as we will show, that long and short in the way the Kimchis suggested were unknown in Hebrew. What is more, when grammar suggests that vowels lengthen or shorten, they do not necessarily change into the ones that Kimchis think they ought to change into. Furthermore, why were totally different signs used for long and short vowels in some cases, and not in others? To get a better idea of the origin of the long and short vowels in the way they are presented here, we might take a look at the vowels in Latin and Greek. Greek and Latin Vowels Greek and Latin poetry is characterised by metre where there is a rhythmic sequence of long and short syllables. (Don t worry too much if you do not quite understand this. You ll soon see what we re getting at.) Sometimes two short syllables can be replaced by one long one or vice versa. Where the accent of the word occurs (which is important in the metre of English poetry) does not matter. What is a long or short syllable in Greek or Latin? A long syllable is one that has either a long vowel, or a short vowel that is followed by two consonants (roughly). Otherwise the syllable is short. The relevant point for us is that by understanding the metre and so on, one can tell sooner or later whether a vowel in a certain word is long or short. Now consider the vowels in Latin. The Romans wrote five vowels I E A O V, which we will represent by i e a o u, each of which can be long or short. In French and Spanish, both derived from Latin, vowels are short. The Kimchi family moved from France to Spain, and in those days every educated person was familiar with Latin. So it is not difficult to guess where the idea came from. Greek is slightly different. Again there are the ten vowels, five long and five short. (The Greek u is pronounced in a different way, but that does not affect the issue.) However, unlike Latin which has five letters for the five vowels, each of which can be long or short, Greek has seven! For the i, a and u sounds they used the letters ι α and υ which could each be long or short. But for the e and o sounds they used different letters for long and short. ε and ο were always short, whereas η and ω were always long. So in Greek there are not five, but seven letters, representing the ten sounds.

15 Vowels in Classical Hebrew 15 The Greeks took their alphabet from the Phoenicians and adapted it for their own use. Why then did they not either use just five vowel-letters (as in Latin) or take ten vowel-letters? The conclusion we may draw from this (which has an important relevance in Hebrew) is that since they did not inherit an alphabet with vowel-signs but devised it to suit their needs, the long and short i, a and u sounds were sufficiently alike not to matter, whereas with the e and o sounds the long and short vowels differed not only in length but also in the nature of their sound. Now look back at the original list we gave of the seven Tiberian vowels and you will see a parallel. There are two sounds for e, two for o, but only one each for i, a and u. This does not imply that Greek influenced the Hebrew, or vice-versa, but merely that the same vowel distinctions occurred, and that just as the Greeks did not consider actual long and short alone as a criterion for writing a different letter, so the Tiberians did not use different symbols to write long and short vowels. If there were long and short vowels (and we will see that there possibly were), the difference was slight and considered unimportant, it did not affect the vowel sound and did not merit using a different symbol. Certainly it did not merit inserting (although this would have been not by the Masoretes but by the scribes who wrote the books) an obligatory extra vav or yod to indicate length, and omitting it to indicate shortness, as Kimchi suggested.

16 16 Vowels in Classical Hebrew Long Vowels and Short Vowels Depending on which language you are considering, there are three totally different concepts in the length of vowels, which the Kimchi system confuses. We need to examine them carefully. Concept 1. The short vowel has a particular sound. The long vowel is the identical sound, lengthened. We rarely meet this in English, but consider then and there. Concept 2. The short vowel has a particular sound. The long vowel is similar, but not quite the same, and longer. This we do meet in English, though our longer vowels tend to be impure and often turn into diphthongs, but not always. Compare carefully for instance the vowel sounds of the English foot and boot, or bit and beat, or the letter e in them and they. German has better examples, e.g. hat and haben. In each case there is a short vowel and a corresponding long vowel, but the two differ in sound quality, not just in length. Concept 3. All the vowels in a particular language can be arranged in order of length this is irrespective of the sound of the vowel, some being inherently longer than others (and it depends on the language). This is relevant in Hebrew, as we shall see. [It is perhaps worth disposing of something else. We were taught in school that there are long and short vowels in French, for example two a sounds, one like the English u in bus and one like the a in father. This is true as far as it goes, but the latter is like the a in father only shortened. In fact there are no long vowels in French, all vowels are short. So the following comment does not apply to languages like French.] The relevance of all this is when in a given situation a vowel needs to be lengthened or shortened. For example, a longer-sounding vowel may need to be shortened when the word itself is lengthened and the accent is moved to another syllable, or vice-versa. The whole Kimchi system is based on Concept 2. Were it based on Concept 1 it would be hard to justify a totally different symbol for e and ē, or o and ō. At this point it is worth our while to consider the Babylonian system.

17 The Babylonian System Vowels in Classical Hebrew 17 It is not our purpose here to describe the Babylonian system in full, or even in brief, but merely to point out certain aspects of it that throw light on the Tiberian system. The Babylonian system makes fine distinctions that the Tiberian does not, although in one case the reverse is true. Leaving aside the sheva, the Tiberian system has seven vowels, plus the three hatafs at most ten. The Babylonian system does not have the hatafs as such, and only has six vowels (no segol), but each of these has three different forms, so that in all you should have eighteen different vowels. [In practice there are only sixteen, because one of these only has one form, but that is a technical detail.] What are the three forms? A. The standard form of the vowel (placed above the letter). Used when the vowel is accented (even with a secondary accent), or in an open syllable. (If it is accented it does not matter whether the syllable is closed, i.e. ends in a consonant, or open, i.e. the vowel ends the syllable.) B. The vowel with a short horizontal line above it. Used when the vowel is in a syllable closed by a strong dagesh, but not accented. C. The vowel with a short horizontal line below it. Used when the vowel is in a syllable closed by another consonant, but not accented, and also in cases corresponding to a hataf in the Tiberian system. Why the three forms? Clearly the length of the vowel is different in each case, but it is almost certainly long or short according to Concept 1 above, and not according to Concept 2. Such a difference we will examine it in a moment is natural in most languages, and it is reasonable to assume that it existed in the Tiberian pronunciation as well, only that the Tiberian Masoretes did not consider it worth worrying about. For all that, it has a bearing on grammar. It is only natural to lengthen a vowel slightly when it is in an open syllable or when it is accented that is to say, some find it natural to lengthen it, but nobody would shorten it in such a case. We may therefore assume that A above is the longest. We know that in the Tiberian system the hataf is something half way between a full vowel and a sheva, so that which corresponds to it, namely C above, is the shortest. This means that a vowel in a syllable closed by a strong dagesh (B) is a bit longer than one in a syllable closed by a single consonant. Essentially, A is longer than B, which is longer than C.

18 18 Vowels in Classical Hebrew We mentioned above that one of the Babylonian vowels (it is the holom) has only one form. It is always A (i.e. long), and if it is expected in a situation where a shorter form (B or C) is required, then it changes, that is to say a different vowel is substituted where the Tiberian, in the same text, has a holom. This shows that a vowel tends to lengthen but only according to Concept 1 above when it is accented or in an open syllable. Let us bear this in mind. Return to the Tiberian If we postulate that something similar takes place in the Tiberian pronunciation, we see that we can immediately discard the rule that we were taught that a long vowel in a closed syllable must take the accent. What happens is that both in an open syllable and in a closed accented syllable, the vowel tends to lengthen, whereas in a closed unaccented syllable the vowel tends to shorten. The length of the vowel does not determine the position of the accent; the position of the accent and the nature of the syllable help to determine the length of the vowel. This makes a difference to grammar, because grammar may determine the position of the accent, which in turn determines the length of the vowel, not the other way round. However, if we wish to apply what we have learned from the Babylonian system to the Tiberian we must take some other things into account. In the Babylonian, the holom can only be long in the new sense, that of Concept 1 and not of Concept 2. The segol does not exist. All the other vowels (including patach and tserey) can be long, short or medium. In the Tiberian, something similar occurs, but different in detail. In the Tiberian, the holom normally can only be long (but the Aleppo Codex has shown some exceptions), and so too can the tserey. The segol can only be short, and the patach as a rule likewise, though the behaviour of the latter is a bit more complicated. The hirik, the shuruk/kubbuts, and the kamets can all be long or short but not in the way Kimchi suggests. A shuruk or hirik in an open syllable tends to be pronounced slightly longer than in a closed unaccented one. There is one interesting parallel with the medium form used in a syllable closed by a strong dagesh in the Babylonian system. In a closed unaccented syllable, kubbuts and kamets are generally interchangeable, but where the syllable is closed by a dagesh the kubbuts is preferred, apparently it is slightly longer (on Concept 3, see below). This is noticed in the pual form of the verb (which normally has a dagesh preceded by a kubbuts, but in four-letter roots where there is no dagesh the kamets is often used), and more so in the hophal, which before a dagesh is almost invariably huphal.

19 Lengthening and Shortening of Vowels Vowels in Classical Hebrew 19 We then come to Concept 3. It is possible to arrange the vowels in order, such that one vowel is inherently longer or shorter than its neighbour. Very often there is a tendency to lengthen or shorten a vowel, and this is done by moving along the scale, i.e. by changing it to another nearby which is inherently longer or shorter. This is not lengthening (or shortening) the same vowel (Concept 1), nor is it changing it to the corresponding (but very slightly different) long or short vowel (Concept 2), but changing it to a different vowel that is inherently longer (or shorter). In Hebrew this is done very often, and not only when grammar demands a longer or shorter vowel, but often when speech makes it preferable, and in such cases it is very often difficult to lay down rules. The best that can be done is to lay down guidelines for tendencies, with no explanation required for exceptions or inconsistencies. As simple examples, in an open syllable a tserey is shorter than a hirik. Of course a hirik in a closed unaccented syllable is shorter still, but the tserey in that case cannot be compared, because it never exists in a closed unaccented syllable it would change to something else. Likewise in an open syllable a holem is shorter than a shuruk. The Effect of the Missing Dagesh The letters + ( 6 $ and nearly always < cannot take a strong dagesh, they cannot be doubled. (In the four cases where w contains a dot this is not a dagesh but a mapik, which does not double the letter, but merely emphasises that the$is not silent.) Instead we are told that the previous vowel lengthens. Not only does it in many cases remain unchanged, but often the reverse it shortens! What happens is this. The vowel followed by a dagesh is in a closed syllable. If the dagesh disappears (because the letter cannot be doubled) the vowel finds itself in an open syllable, and tends, though remaining unchanged in quality, to lengthen according to Concept 1. However, it may then shorten according to Concept 3. For example ¼-HºD/KP<Hx becomes simply ¼-HºD/K<Hx (the vowel, finding itself in an open syllable, lengthens a bit according to Concept 1) which then shortens to ¼-HºD/K<Ix (because on Concept 3, in an open syllable a tserey is shorter than a hirik). The same thing happens at the end of a word, where, likewise, a strong dagesh is not allowed (except in a º and not always then). We have ¼-H O+, but when the vowel ends the word, it becomes ¼;O+ (or ; + which is the same thing there is no

20 20 Vowels in Classical Hebrew reason for inserting a )). As this is now in an accented syllable, it would tend, while remaining unchanged in quality, to lengthen according to Concept 1. However, it then shortens to ¼;N+ (because on Concept 3, in an accented syllable a holem is shorter than a kubbuts/shuruk). We do not see this, because we have been indoctrinated with the idea that there would be a difference between ¼; O+ and ; +. Likewise though we have ¼- Hx H0, without the suffix ¼% H0 shortens to ¼% I0. This shortening does not always happen. When we find the piel form ¼ I+H it is not that H has failed to lengthen to I, but that it has failed to shorten toi. The rules of grammar must be expressed differently. Mediaeval Hebrew Poetry The main requirements of modern (other than so-called contemporary) English poetry are that it should rhyme and scan. Also the accent should appear correctly in the scanning. In Biblical poetry, scanning seems to be unknown, and rhyming occurs in only two poems Lemekh s dirge, and the bards song deriding Moab and praising Sihon. However, there were other characteristics such as parallelism. Greek and Latin poetry demanded scansion, on which it was very strict, but the accent was not taken into account. They didn t bother about rhyming. The mediaeval Jewish poets (authors of the piyyutim ) tried to do everything. Their poetry often had to rhyme, scan and fit in with acrostics. Scanning was the weakest point, if anything had to be sacrificed. However, when they wrote songs, they were more careful, and often let the acrostic go by the board. When it came to scanning, they tried to introduce metres, but not having long and short vowels (before the Kimchis invented them) they had problems with their syllables. The solution they found was interesting. Instead of using short and long syllables depending on the nature of the vowel etc., as the Romans and Greeks did, they used simple and compound syllables; these could be open or closed, it made no difference. A compound syllable is one that starts with a consonant and a sheva or hataf, followed by another consonant and a full vowel (with or without a closing consonant). A simple syllable just has a consonant followed by a vowel (with or without a closing consonant). They then composed using compound syllables where the Classicals used long ones, and simple where they used short ones. Or, if you prefer it, a short syllable contained one (full) vowel and a long syllable contained one-and-a-half vowels, the half coming first. [* I am grateful to the late Mr. Meir Brachfeld for showing me this.]

21 Vowels in Classical Hebrew 21 Two wonderful examples of this are the hymns that we sing at the end of the services on Friday nights and Sabbath mornings, viz. Yigdal and Adon Olam respectively. Yigdal, from left to right, goes Adon Olam, left to right, goes Watch it now, in Hebrew, from right to left. (We will take the second line of Yigdal where the Divine Name does not occur.) The long syllables are in large type, the short ones in smaller type ¼}¼' ¼ ¼+ H-D ¼'¼-H+ L- ¼3¼-I$D) ¼'L+ J$ ¼}¼> ¼ ¼z D+K$D0 ¼7¼}¼5 ¼3¼-I$ ¼1K&D) ¼1L0 D6J4 For the other example ¼$L< D%H4 ¼<¼-H:D- ¼0L ¼1J< J,Dx ¼ K0 L2 ¼<J F$ ¼1L0 ¼}¼6 ¼3¼}¼'F$ Where this goes wrong in the song, either the poet was taking a liberty, or (more likely) the original text has changed. An interesting example is the last two lines of Yigdal in the Sephardi version, which do not occur in the Ashkenazi version. Did the Sephardim add them to the original, or did the Ashkenazim drop them? The fact that these two lines in no way fit the metre shows that they were added, not by the original author but by someone who did not understand the metre at all. However, the purpose of this section is not to discuss mediaeval poetry, but to show that the poets who wanted to imitate Latin and Greek styles could not do so properly, because Hebrew did not have long and short vowels the way Latin and Greek did. At least, not until the Kimchis invented them. Primitive Scribal Customs When considering the effects of the Scribes on the spelling of written words (including and excluding vowels), we must consider two distinct groups of scribes: those who wrote before the Masoretes introduced vowel signs (whose customs one may be tempted to call primitive scribal customs ), and those who wrote after. At all events, Hebrew scribes are notorious for their laziness. This first became evident

22 22 Vowels in Classical Hebrew when they decided that if a letter was doubled in pronunciation, it was only necessary to write it once, thus later necessitating the strong dagesh. But the worst outrage they committed against the Hebrew language, in which they were followed by authors and later by printers, was the introduction of abbreviations. These, according to Rabbi Judah, were (as we all know) first inflicted on the Egyptians along with the Ten Plagues, but later they were inflicted on us all, and there is no doubt that today abbreviations are the curse of the Hebrew language far, far worse even than the now obsolete conversive vav! However, it is not these matters that concern us, but certain random usages along with a few conventions that were not often kept, on the part of the early scribes; and the way the later scribes took the masoretic vowel system and in the interest purely of sheer laziness destroyed its logic, creating confusion that makes grammar so difficult. Two problems faced the early scribes, with regard to spelling. The first was the use or non-use of a silent letter to indicate the previous vowel. The tendency was to preserve a letter that had originally appeared in the root, but later became silent, and this practice was kept up fairly consistently with the letter$. There are cases where it was omitted, but these are rare; also cases where it was inserted, but these are even more rare. More complicated were the numerous cases where the letter vav weakened, and instead the previous vowel was replaced by o or u, the vav becoming silent; or where the letter yod weakened, and instead the previous vowel was replaced by i or é, the yod becoming silent. Here, as the silent letter originally belonged to the root, it should have been left in, but this was not always done. In fact it was very often not done, so that one could no longer rely on it. People no longer assumed that the silent letter referred to the original root, but merely helped in indicating the previous vowel. At the same time, since it performed this function, the silent letter (vav or yod) was often included even where it had not previously been part of the root, to indicate the previous vowel in the same way. Thus in the course of time the silent letters tended to simply indicate the previous vowel, whether or not they were part of the original root. The question was when to use them and when not to. Discovery of the Dead Sea Scrolls has shown that customs varied even with books of the Bible. Two scrolls of Isaiah have been found, one of which uses these silent letters extensively, while the other is quite frugal with them. In this respect even the Talmud admits that we cannot be sure of the original spelling, but the version we have is a standardised version determined by the Masoretes. Here we find that the usual conventions for when to use these silent letters and when not to

23 Vowels in Classical Hebrew 23 varies considerably from book to book within the Bible, and in every case there are many exceptions. One convention is a tendency not to put two silent letters in one word, so that where two are apparently called for they will put one or the other sometimes in the same word, one in one place and the other elsewhere. For a full consideration of the subject, with some sort of guideline rules, the reader must refer to ><)52( ><)52 by Eliahu Bahur (Elias Levitas) written about five hundred years ago. Here is not the place to enter into details, our purpose is merely to give the reader an idea of how complicated it is. The need for this will become apparent. One convention however was upheld to a far greater extent. The silent letter was not usually put in after a vowel in an unaccented closed syllable. In such a case the vowel would be short. (Later on, in the Talmud for instance, the silent letter was indeed often put in after a vowel in a syllable that was closed by a dagesh, and it is often done today, but that does not concern Classical Hebrew.) On the other hand it was often (but by no means always) put in after an accented vowel, or a vowel in an open syllable, where as we explained earlier there is a natural tendency to lengthen the vowel somewhat. This practice became more common in the later Biblical books, it is less common in the Torah. However, it led many grammarians to begin to believe, and Rabbi Joseph Kimchi to openly state, that the vowel not followed by the silent letter (other than kamets which we are not considering here) is short and that followed by a silent letter is long; and conversely that a short i or e should not be followed by a silent letter, and a long or ē always should, unless there is a good reason to the contrary (which we may or may not be able to discover). There are in the Torah alone hundreds of exceptions that make utter nonsense of this rule. The second problem followed a desire, for some unknown reason, to always conclude a word in writing with a consonant which does not have a vowel after it. Where the word already ended in a consonant sound or with a silent letter, there was no problem. Otherwise they would usually add a silent letter after i or é a yod, after ó or u a vav, and after a kamets or segol (or a tserey derived from a segol) a silent(. This covers the vast majority of cases, but they were not always consistent and there are exceptions, some of which can be grouped in sub-rules, but many of which cannot. A major group of exceptions is in the personal endings > (3)4 ( (.)/ where these are followed by a kamets sound, where they did not usually bother to add a silent(. (After(they never did, after (3)4 they mostly did, after (.)/ and > they seldom did, except in the word (>$ you masc. sing.) Of course it is more complicated, but this gives you the general idea. This custom greatly influenced the later scribes when they messed up the vowel system.

24 24 Vowels in Classical Hebrew [They also inserted a silent yod to indicate when a noun with a suffix was meant to be plural, regardless of what vowel preceded it, e.g. ¼ ¼4¼-I4Lx ¼)¼-L4Lx ¼ª¼-J4Lx ¼1¼-H4Lx. Occasionally they forgot to do this, but were more careful not to do the reverse.] Masoretic Conventions It is necessary to continue with all this, in order to see how the eventual messing up of the masoretic system by the scribes misled the grammarians. The Babylonian Masoretes introduced vowel dots and dashes placed above the letters to represent the six vowel sounds. The Tiberian Masoretes introduced vowel dots and dashes placed below the letters to represent the seven vowel signs, except for the holem which was a dot placed above the letter. (Only later was it moved over to the left.) Where the vowel was u followed by a silent vav, they left out the vowel and instead put a dot in the middle of the vav thus. However, the Babylonian scribes were consistent. If a vowel was followed by a silent vav or yod, they attached the vowel to the silent letter, instead of to the letter to which it belonged. (See the appendix.) The Tiberians did this with the vav, so that if it followed a holem you had }, but not with the yod. This led people to believe that the } is itself the vowel! Later, as explained above, people began to think that the silent yod is also, if not the vowel, then an integral part of it, instead of being merely an indicator of a preceding vowel. The Masoretes placed a vowel symbol under every letter. It could be a full vowel, or a sounded sheva, or a half-vowel (hataf), and if there was no vowel sounded they place a silent sheva there which symbolised a non-vowel. But even the non-vowel sign could be regarded as a vowel symbol. The exception was a silent letter, whether in the middle of a word or at the end, which had no vowel sign underneath, not even a silent sheva to show that the consonant was pronounced without a vowel the absence of a symbol showed that the consonant was not pronounced. All this is logical. Scribal Scribble As a slight digression, not directly connected with vowels, the Masoretes also used three other conventions. They put a dot in a letter to indicate that it was to be doubled in pronunciation, and a line above it to show that it was not to be so doubled where you might think it was. This was maintained. Then they put a line above a silent(to show that it was silent this was at the end of a word, so as to