Kernerman Dictionary News • Number 12 • July 2004

Milon ha-Hoveh and Milon Sapir

Mordechay Mishor

 Mordechay Mishor was born in Argentina in 1939. He graduated from the Hebrew (Shazar) College in Buenos Aires, immigrated to Israel in 1963, and studied Bible and Hebrew at the Hebrew University of Jerusalem. His thesis “The Tense System in Tannaitic Hebrew” was completed in 1983, under the supervision of Ze'ev Ben-Hayyim. In 1967 he began to work in the Department for Ancient Literature in the Historical Dictionary Project of the Academy of the Hebrew Language, and in 1989 became the department head. Since 1985 he has been coordinating the work on the Dead Sea Scrolls, the epigraphic documents, and the Talmud and Midrash literature, and since 1998 he is a member of the Academy’s Grammar Committee. He has taught at the Negev (Ben Gurion) University and the Hebrew University, and has cooperation with universities in Spain. Dr Mishor has written and lectured extensively in Israel and abroad. His main research interests include Mishnaic Hebrew in all its linguistic and philological aspects, as well as the living reading traditions of Jewish communities, and the relationship between the Hebrew traditions in the Middle Ages and the language revealed by the epigraphic findings.

This article was translated from Hebrew by IJK, translation edited by Raphael Gefen.
General cases of using he/him/his refer to both masculine and feminine.
The Hebrew version:

In this short article two dictionaries are reviewed: Milon ha-Hoveh [Dictionary of the Present (MH)] and Milon Sapir [Sapphire Dictionary (MS)]. My aim is to present these dictionaries in the light of the lexicographic principles that guided their editing, and not to criticize the end product for minor faults or random slips.


Milon ha-Hoveh

The distinctiveness of MH and those following in its suit is in representing the verb in the present tense form (the present participle). The idea was conceived in the course of my work in the Historical Dictionary Project of the Academy of the Hebrew Language, and was made public in 1985 at the Hebrew University, to an audience of linguists, in a programmatic lecture where I dwelled upon the qualities needed from a practical dictionary.1

Representation of the verb in the present tense form is based on the cognizance that the present participle in Hebrew fulfills a double function, of a noun (as a substantive, an adjective or an adverb) and of a verb. In traditional dictionaries the participle in its nominal meaning will appear according to the first letter, and in its verbal meaning it will be represented by the verb, and appear according to the first letter of its root. Thus, for example, the word menahel would appear under the letter Nun [N] as a verb in the present tense (the root NHL [to manage, direct]), and under the letter Mem [M]  as the name of a professional (the entry menahel [director, manager]). However, this recognition demands consideration, and the distinction between the two usages is not always sharp enough, such as in the phrase “so-and-so menahel a factory” [manages / is the manager of]. Listing the verb in the dictionary in the participle form exempts the user not only from the vacillation towards making a decision, but from the sheer awareness of the problem. In this aspect MH is particularly friendly to those whose linguistic knowledge is not professional, or is not professional enough, but is practical-functional. That is, the less aware the user is of the double function of the participle, the more suitable MH is for him. The constant reminder to the user, to look for the verb in the participle form (identical to the present tense), is implied in the dictionary’s name – Milon ha-Hoveh.

In that programmatic lecture, besides making the suggestion about the place of the verb in a practical dictionary, there were a number of recommendations. The last paragraph in the lecture sums up the standpoint that was subsequently taken in MH: “I will end with some recommendations, considered self-evident: A practical dictionary should be written with the usual spelling, which is the ‘plene’ spelling (including the headwords). In a practical dictionary there is no need for etymology, nor attribution to historical layers; on the other hand, one should be generous with stylistic and normative evaluations. Excerpts are superfluous. Entries that are not used in our present-day contexts are superfluous. If we add to this the cancellation of redundancies that are created by distinguishing the nominal participle from the verb – we would gain another merit: that the dictionary will be short.

The work on MH began in autumn 1988. During the editing many problems rose or became acute, and the solutions were directed along one guiding principle: to help the user find the requested word quickly, to define it in short and clearly, and to avoid excessive information that might distract his mind from the text in front of him. (The search speed in this dictionary as compared to others was proved in a survey carried out by the publishing house in the early editorial stages in order to assess its commercial worthwhileness.

As recommended in the above citation from my lecture, the entire dictionary is written in plene spelling (the “standard non-vocalized orthography”), with the addition of auxiliary diacritical vowel marks when needed, but the vocalized (‘defective’) spelling, also called ‘grammatical’ spelling, is attached to each headword. There are references from the grammatical to the plene spelling, in accordance with which the dictionary was compiled.

A radical morphemic principle was adopted in editing the dictionary for the separation of homonyms: these were separated into different entries only if their root or inflection varied. For example, according to this principle the word musar was divided to two entries, one from the root YSR (“hatafat musar” [moralizing], “musar klaiot” [remorse]) and the other from the root SWR (passive of “mesir” [remove]); along this principle the two meanings of the verb “no'el” [putting (shoes) on and locking (a door)] were given in one entry, and “qeren” was divided to different entries according its plural: “qranot” [horn musical instruments, funds], “qarnayim” [rays of light, animal’s horns]. Participial forms with two usages, nominal and verbal, were put under a single entry, and the different usages were indicated in the definition according to their grammatical behaviour, also for nouns used in adjectival or adverbial functions. Adverbs that are constructed on the base of a noun with a preposition (“le-'olam” [for ever], “bi-frat” [in particular]) were given as sub-entries. Adjectives that are derived automatically from substantives by adding the suffix –y, as well as abstract nouns likewise derived with the addition –ut, were given as sub-entries. This approach broadens the definition of the actual entries, but reduces considerably the number of entries and accords the dictionary a “compact” character.

The dictionary consists of the information that is indispensible for the reader of present-day texts. In this respect the dictionary is singled out precisely by what was decided not to include in it. The planned vocabulary was based on 20,000 main entries, said to represent the most frequent words, in their common meaning in modern Hebrew. As compared with other dictionaries, which allocated to the nominal participle its own entry, which separate the noun usages according to their functions (a substantive, an adjective, an adverb), which tend to separate the homonyms according to a semantic or etymological principle and not a formal one, and which bring the adverbial fuction of prepositional phrases as entries on their own – compared with these dictionaries the figure 20,000 in MH is much more comprehensive. In the end we were unable to limit ourselves to this number and it was extended to 21,000 (30.7 main entries per page on average [21,000:684]).

A substantial concession was made in the grammatical information accompanying each entry. Here the principle of predictability was applied, so that only what cannot be known by the actual grammar mechanism was specified. For example, for nouns the gender was indicated and the plural noted (from the form alone it is not possible to know, for example, the gender of the word “'eretz” [land, country, ground], or its plural); the feminine form of the adjectives was also indicated, but not their plural form. The verb had an indication of the government (its depending preposition), enabling the different senses to be differentiated. In contrast, the part of speech of the entries was not indicated, nor was the conjugation stem (binyan) of verbs; and these are bold innovations. Nor was the historical layer that is attributed to the word indicated, since this is professional information that does not contribute to the word’s meaning or stylistic status. On the other hand, the root of each entry was added (as far as was known or possible), this being the semantic foundation relating a word to its “family” members. The data was carefully filtered, to release the user from being overburdened. Incidentally, the fact that the dictionary is short is an end result of this principle, not an aim in itself, because when the editors found it necessary they did not hesitate to expand (see below re the verb conjugation key).

Examples of usage were given only when the definition alone was not enough to make the usage of the word clear. In principle, this dictionary was not meant to teach how words are to be used (“how do you say …?”), but to provide their interpretation (“what does … mean?”), after the user has come across them in their natural context in conversation or text.

A lot of effort was invested in grammatical help for locating the entry being sought. A key was appended for this purpose including all the past, future and infinitive forms of the verbs in the dictionary, with referral to the dictionary entry (e.g. “leishev > yoshev” [to sit > sitting]). This appendix contains 78 pages. Another appendix, a key of the roots and their attributed entries (37 pages), is meant to help those interested in revealing the meaning of a word that for some reason was not included in the dictionary, according to its “family ascription”.

The official standard was set as the point of reference for marking the stylistic and normative status of the entries, that is, the standard of the Academy of the Hebrew Language. This is expressed not by censoring non-standard entries, but by “grading”: literary, popular, vulgar, slang. The fidelity of MH to the Academy settings in all domains – the spelling, the formation, the usage, the relation to foreign words, etc. – has made this dictionary an authoritative aid, which has found its place on the desks of writers and editors.

My work on MH was done while working on the Historical Dictionary Project. Shoshana Bahat was then the scientific secretary of the Academy, until her retirement in 1990. MH appeared on 1 February 1995. The last stage of our work was very intensive. During that time Shoshana Bahat fell ill, and the final crafting was cast on my shoulders. Shoshana was not fortunate enough to see the dictionary published (she died in November 1994). While still active she managed to accompany the first steps of a large-scope dictionary, edited in the method of MH (subsequently Milon Sapir), as linguistic advisor, and confided in me her quandries concerning her new interest. I completed the work on MH physically and mentally exhausted, and my sole desire was to return and invest my full vigour in the Academy’s Historical Dictionary, which had naturally shifted for a while from the centre of my activity.


Milon Sapir

The success pronosticated by the publisher for Milon ha-Hoveh long before it appeared gave way in 1991 to the idea of publishing a series of dictionaries of varied scope along this dictionary’s editing principles (“ha-Hoveh method”). The dictionaries compiled according to this method are called Milonei Sapir [Sapphire Dictionaries] 3. Thanks to their marketing momentum, ha-Hoveh method has had a wide dissemination. I will discuss here only the dictionary called simply Milon Sapir (MS), or Milon Sapir ha-Merukaz in its full name [The Concise Sapphire Dictionary].

The substansive difference between MS and MH is in the extent. Yet that is not the only difference, and I will devote the rest of my words to this.

The full title of MS announces that it was edited in ha-Hoveh method. In fact, the dictionary is built on the basis of MH itself, but the entries were expanded, with the addition of, naturally, many main entries and sub-entries, and other changes introduced in the setting of entries.

The editing principles are not different from those of MH: MS has plene spelling, and the entries include the vocalized form in the grammatical spelling. The auxiliary diacritical marks were added throughout systematically (unlike in MH). The verb is represented, of course, in the participle, and the government was indicated. The different senses of the entry are numbered (in MH they are separated by a semi-colon). The roots with their derivatives are integrated in the dictionary core (in MH – in an appendix). As with MH, the headwords have no examples of usage.

Among the bold innovations of MH, the renunciation of the verb conjugation stem was adopted. The formal principle in dividing homonyms was also kept. In contrast, an indication of the part-of-speech was added to each entry (verb, noun, etc), as well as the historical layer ascribed to each meaning. The most striking concession, and most deplorable in my view, is dropping the verb forms key that supplemented MH.

The main contribution of MS as compared with MH is, as above-mentioned, in its number of entries. As it says on the cover, the number of entries is 90,000 (81.2 on average per page [90,000:1,108]; actually, the dictionary has not a single page that contains such a vast number of entries). However, it is not the precise number of entries in the dictionary that interests us in this review, but the principle practised in expanding the vocabulary. So, in addition to new entries and idioms that were not in MH, whose due place in a dictionary like MS is in no doubt, one domain was expanded to encyclopedic scope, that of geography: “… all the countries, capitals, biggest lakes and seas in the world, tallest falls and mountains, and special sites, such as tunnels. With regard to Israel all the settlements with a population of five thousand and over have been included.”
4. Selecting this domain for broadening a dictionary which is not encyclopedic seems somewhat strange.

I have mentioned that the part of speech was added in each headword in MS. Following it came the words’ various functions, according to the parts of speech, as independent entries. For example, the entry qasheh [hard, difficult] was divided to three: 1. verb, 2. adjective, 3. adverb (but the adverb qashot [severely] was included for some reason in the adjective). The same applies to names that are not in the participle, for example, the entry yaqran [person charging exorbitant prices] was split in two: 1. adjective, 2. substantive. Unlike in MH, adjectives with the suffix –y and abstract nouns with the suffix –ut were all given as entries on their own. Thus, for example, instead of the main entry mu'amad [candidate / erected] in MH, which includes also the abstract noun mu'amadut [candidacy] as a sub-entry, MS has four entries: 1. mu'amad verb, 2. mu'amad adjective, 3. mu'amad substantive, 4. mu'amadut. In MH the entry mu'amad began with the verbal meaning (passive of erects), and then stated that in the present form there is another meaning (candidate); there was no note about the distinction between the substantive and the adjective, since both have identical grammatical categories: feminine and plural. In comparison with MH, MS has a simulated broadening of entries, which stems from a different editorial policy.

As in MH there are references from the entry in grammatical (defective) spelling to the entry in non-vocalized (plene) spelling. MS has the addition, systematically throughout the dictionary, of references from the past form of the verb to their present form. This is conspicuous in the letter H, particularly with the beginning hu–, where some pages contain almost only references. In fact, the sub-title of the dictionary – “according to ha-Hoveh method” – makes these references superfluous.

As against the compact MH, MS looks “respectable”, impressive in its dimensions, and more generous with grammatical and historical information. The reduction in this information, which was perceived by the editors of MH as a constructive breakthrough – if such an oxymoron is permitted – of the fences of Hebrew lexicographic tradition, seemed a too far-reaching dare to the editors of MS, who toed the line with “conventional” lexicography. Anyway, MS proves that ha-Hoveh method is applicable in dictionaries with varied scopes and/or of a different character, as is clear from the range of Sapphire Dictionaries published since then.



1.      The lecture ‘What Distinguishes a Scientific Dictionary from a Practical Dictionary?’ was given on the anniversary of the Department of the Hebrew Language at the Hebrew University (1985), and published in Balshanut Hofshit, 24.69-74, 1986.

2.      The essence of the survey results appeared in a letter by Ilana Shkedi from Sifriyat Maariv, published in Leshonenu La'am, 47.43, 1996.

3.      Including Sapphire Encyclopedic Dictionary (7 vol. 1998; 3 vol. 2002), Sapphire School Dictionary (1999), Word by Word (2000), First Tongue (2000).

4.      From the Preface.


Milon ha-Hoveh
a Practical Dictionary for Standard Hebrew
S. Bahat and M. Mishor
Eitav Publishers – Sifriyat Maariv, Tel Aviv, 1995

Milon Sapir
Concise Hebrew-Hebrew Dictionary according to ha-Hoveh Method

Chief Editor: Eitan Avneyon
Linguistic Advisor: Shoshana Bahat
Scientific Advisor: Raphael Nir
Scientific Proofreading: Yitzhak Shlesinger
Hed Artzi Publishers and Eitav Publishers, Tel Aviv, 1997
1108 pages (including roots of associated entries) + introduction

K Dictionaries Ltd
10 Nahum Street, Tel Aviv 63503 Israel
tel: 972-3-5468102 • fax: 972-3-5468103