The representation of spoken Hebrew in Eretz-Israeli fiction films from the 1930s

This article examines whether the dialogue in the first two full-length Hebrew narrative films reflects spoken Hebrew of the time. The first part examines phenomena that, according to written testimonies in grammar books and prescriptivist writings, were typical of spoken Hebrew during the British Mandate period. It shows that these phenomena were not represented in the films from the 1930s, whereas in later films there is a significant rise in their frequency. The second part examines phenomena that are not typical of spoken speech today but are represented in the dialogue of the films from the 1930s. In later films they become very rare or vanish altogether. An analysis of the findings shows that the film-makers did not regard it as the function of cinematic dialogue to authentically reflect spoken speech, but rather adopted the normative approach to Hebrew, which strongly preferred high-register Hebrew over authentic usage.

The representation of spoken Hebrew in Eretz-Israeli fiction films from the 1930s M i r i B a r -Z i v l e v y h e B r e w u n i v e r s i t y, j e r u s a l e M , i s r a e l a B s t r ac t This article examines whether the dialogue in the first two full-length Hebrew narrative films reflects spoken Hebrew of the time. The first part examines phenomena that, according to written testimonies in grammar books and prescriptivist writings, were typical of spoken Hebrew during the British Mandate period. It shows that these phenomena were not represented in the films from the 1930s, whereas in later films there is a significant rise in their frequency. The second part examines phenomena that are not typical of spoken speech today but are represented in the dialogue of the films from the 1930s. In later films they become very rare or vanish altogether. An analysis of the findings shows that the film-makers did not regard it as the function of cinematic dialogue to authentically reflect spoken speech, but rather adopted the normative approach to Hebrew, which strongly preferred high-register Hebrew over authentic usage.
T h e f i r s t h e Br e w-s p e a k i ng fiction films were produced in the 1930s. They provide some of the earliest examples of recorded Hebrew dialogue. This article seeks to determine to what extent the dialogue in these films faithfully reflects the living spoken language that was evolving in the Yishuv 1 at the time, and whether there is a difference in this sense between a silent film and a talking film produced during this period. This article is based on my Ph.D. dissertation, supervised by Yael Reshef. I would like to express my deepest thanks to her. I wish to thank Vera Agranovsky, Chanan Ariel, Elitzur Bar-Asher Siegal, Edit Doron, Shifra Witman and the anonymous reviewer. Each of them helped me in a different and crucial way with the writing of this article. Support from the Mandel Scholion Interdisciplinary Research Center in the Humanities and Jewish Studies at the Hebrew University of Jerusalem is gratefully acknowledged. The research leading to these results has also received funding from the European Research Council under the European Union's H2020 Framework Programme (H2020/2014-2020)/ERC grant agreement no. 741360, principal investigator Edit Doron. suspect him of stealing a goat. The Bedouins imprison the tourist in a tent, unaware that Oded is watching them from a distance. Oded's friends fly a kite to mark their location and Oded sees it and rejoins them. At this point, the group is also joined by the teacher and a guard. After Oded tells them the tourist has been kidnapped, the guard goes to the Bedouin camp to free him. The Bedouins receive the guard politely and accept his explanation as to the tourist's presence near their camp, but when they come to release the man from the tent they discover he has managed to escape. Oded and the others return to their village, where Oded is received with great joy. The screenplay was written by director Haim Halahmi, who was born in eastern Europe and emigrated to the Land of Israel in 1925. The film's creators and main actors were not native Hebrew speakers, but they lived in the Yishuv for some years before making the film and knew Hebrew well.
Tsabar tells of a group of Jewish pilgrims who experience great suffering and hardship trying to build a settlement in a desolate part of the country. Their Arab neighbours do not accept their presence. A shortage of water and the malice of the Arab sheik lead to a quarrel that culminates in a violent attack on the Jewish settlers. Eventually, the quarrel is settled after the Jewish women dig a well and find water in abundance. Whereas the Jews are presented in the film as rational and proactive, the Arabs are presented as primitive and passive. Olga Ford wrote the script, and Mary Sukhovolsky wrote the dialogue. The actors were mostly from 'Habima' Hebrew theatre. They emigrated to the Yishuv from eastern Europe several years before the film was made. They were not native Hebrew speakers, but they mastered the Hebrew language. The film's director himself did not live in Israel and did not know Hebrew. The film was dubbed in Warsaw by Yiddish theatre actors. The British Mandate authorities censored parts of the film as anti-Arab propaganda, 5 and in the countries around the world where it was screened it provoked intense criticism. Ten years after it was first released, a censored version was produced by Jacob Davidson and screened under the title Halutsim ('Pioneers').
For this study, I used digitized copies of the two movies created by film director Jacob Gross eighty years after their original release. The repaired and digitized copy of Oded ha-Noded was created by Gross using original 5. M. Zimmermann, 'Israeli Films Learn to Speak', Cinematheque 41 (1988), pp. 14-18 (in Hebrew). reels provided by Chaim Halachmi's son. As for Tsabar, Gross reconstructed it from several incomplete copies found in archives around the world.
Part I of this article deals with the question of whether the living spoken Hebrew of the 1930s is reflected in the dialogue in the films from that period. Part II deals with phenomena in the films that are today typical of written (rather than spoken) Hebrew, but whose prevalence in spoken Hebrew of the 1930s cannot be ascertained, due to lack of evidence. These are phenomena that are in a historical process of exclusion seen in the films of the 1950s and onwards. Most of the phenomena discussed in both parts were very scantly represented in the films, but since the two films together have only 575 words of dialogue, the mere presence of these phenomena is significant.

Part I: The representation of phenomena attested to be characteristic of spoken Hebrew during the British Mandate period
As indicated, this part of the article examines whether the films from the 1930s contain examples of linguistic phenomena that, according to testimonies, were common in spoken Hebrew of the Mandate period. To determine which phenomena were typical of colloquial Hebrew in that period, I rely on written testimonies by grammarians, some of them prescriptivists, many of which are quoted in Reshef. 6 These phenomena are not necessarily innovations of MH. Most can be found in earlier layers of Hebrew but were less frequent or were used differently. I focus on phenomena that, according to the grammarians, were common in the 1930s and are also part of the grammar of spoken Hebrew today. For each phenomenon, I will present the evidence for its prevalence in the Hebrew speech of the Mandatory period, and then examine its representation in the films from the 1930s. I will also address its representation in later films. 6

Present participle negated with lo
In contemporary spoken Hebrew, participle forms are generally negated with lo rather than with prescribed 'en -for example, ‫רוצה‬ ‫לא‬ ‫הוא‬ 'he does not want' rather than ‫רוצה‬ ‫אינו‬ ‫הוא‬ -and there is ample evidence this was the case from a very early stage of MH 7 and had some roots in pre-modern eastern European forms of written Hebrew. 8 In Biblical and Rabbinic Hebrew (RH), participles are generally negated with 'en, but negation with lo is found as well, although more commonly in the latter than in the former. 9 Since the rules governing the negation of participles with lo in Classical Hebrew are complex, 10 prescriptivist MH grammarians did not attempt to inculcate this 'correct' usage of the construction. Instead, from the advent of MH, they waged a stubborn war against all uses of lo to negate the participle. 11 Two explanations have been proposed for the shift to the use of lo in this environment. First, the simplicity of the construction, given that lo, unlike 'en, is not marked for agreement and its position in the sentence is fixed. 12 The second explanation is that, due to the influence of foreign languages, the participle in MH is perceived as a purely verbal form, and as such is negated with lo, just like other verbs. 13 The films Oded ha-Noded and Tsabar contain no instances of participles negated with lo. Each includes one example of a participle negated with 'en. In both cases the word used is 'eneni, the Biblical form of 'en in the first- We see, then, that negation of the participle with lo, which was evidently the unmarked option in spoken Hebrew of the time, is not represented in the films from the 1930s. In films from the 1950s this construction begins to appear, alongside participles negated with 'en, and from the 1960s onward participles are negated almost exclusively with lo. 16

Independent subject pronoun before past/future verb
An independent pronoun before past and future verbs -such as ‫אכתוב‬ ‫אני‬ 'I will write', as opposed to just ‫אכתוב‬ lit. 'will write', or ‫כתבתי‬ ‫אני‬ 'I wrote' as opposed to ‫כתבתי‬ lit. 'wrote' -is not a characteristic of Classical Hebrew. According to prescriptivist Hebrew grammarians it is 'superfluous' or 'extra-positional', because the verb in these sentences incorporates an obligatory morpheme denoting the subject, whereas the independent pronoun can be omitted without affecting the grammaticality of the sentence. Hence, prescriptivist grammarians contend that the verb-incorporated morpheme is the subject, whereas an independent pronoun, if present, is an optional element that should only be included for purposes of contrast or emphasis. However, their observations indicate that the use of these pronouns in unmarked contexts was already common in spoken Hebrew of the British Mandate period. Yitzhak Avineri wrote in 1929: 'The pronouns are always added unnecessarily in past and future tense … even in instances that involve no emphasis.' 17 This too was perceived as resulting from the influence of foreign languages. 18 This phenomenon is also found quite frequently in 15. Because the sentence trails off after the negator, it is impossible to confirm whether a participle was actually intended here. Though the existence of participles in the preceding clauses may imply that an additional participle was intended following the negator, this is not necessarily the case -it could have been a noun.
16 Hassidic tales. 19 Observations from the Mandate period mention no differences between past and future tense or between first, second and third person, but data from contemporary spoken Hebrew clearly show that the frequency of the independent pronouns depends on the parameters of tense and person. According to a study by Smadar Cohen, based on a corpus of spoken Hebrew from the early 2000s, 70 per cent of past and future clauses contain no independent subject pronoun. However, when person is taken into account some sharp differences emerge. Clauses in the first person show a clear distinction between past and future: future-tense clauses generally contain an independent subject pronoun, whereas past-tense clauses do not. 20 The films Oded ha-Noded and Tsabar contain no instances of independent subject pronouns in unmarked contexts. In past and future contexts, independent subject pronouns occur only for the purpose of emphasis. The former film includes two such examples, and the latter includes one: 3. ‫מקומכם.‬ ‫את‬ ‫נמלא‬ ‫הנשים,‬ ‫אנו,‬ 'We women will take your place.' (Tsabar) When no emphasis occurs, the independent pronoun is absent:

‫בידינו?‬ ‫קברים‬ ‫לנו‬ ‫נחפור‬ ‫מתי‬ ‫עד‬
'How much longer will [we] dig our graves with our own hands?' (Tsabar) Like the negation of participles with lo, independent subject pronouns are a phenomenon that evidently abounded in spoken Hebrew of the Mandate period, yet it is not represented in the films from the 1930s. In films from the 1950s onward, the pronoun 'ni gradually begins to appear before first-person future-tense verbs. 21

Definite-head construct
In contemporary spoken Hebrew, the definite prefix is frequently appended to the head of a construct -for example, ‫ספר‬ ‫הבית‬ -rather than to the annex -‫הספר‬ ‫בית‬ 'the school' -as prescribed by the rules of formal Hebrew. 22 This phenomenon was common even before the revival of Hebrew as a spoken language, 23 and observations by grammarians indicate that it was already common in the colloquial Hebrew of the Mandate period. 24 The films Oded ha-Noded and Tsabar include only one definite construct each, both instances conforming to the standard rule, with the definite prefix joined to the annex rather than the head: Here is the dance of death.' (Tsabar) Definite-head constructs, then, are not represented in the films of the 1930s although they were common in the spoken language. A corpus of ten films from the 1950s and 1960s also contains no instances of this construction, but later films, from the 1970s and 1980s, exhibit quite a few instances of it. 25

Analytic genitive and accusative constructions
Spoken and written MH differ considerably in the frequency of synthetic versus analytic constructions. Whereas the written language contains many instances of the former, spoken Hebrew prefers the latter. Moreover, in both written and spoken Hebrew there has been a gradual shift from synthetic to analytic constructions. 26 I will elaborate here on three analytic constructions: the 'analytic construct', phrases with independent object pronouns, and phrases with independent possessive pronouns. All three are frowned upon by prescriptivist grammarians, but they are ubiquitous in colloquial MH.
Analytic construct The synthetic construct (involving two nouns without a preposition between them) -for example, ‫ינחבמ‬ ‫םידימלתה‬ 'student exams' -occurs in all layers of Hebrew, including MH. The 'analytic construct', in which the two nouns are connected with the possessive preposition šel -for example, ‫םינחבמה‬ ‫לש‬ ‫םידימלתה‬ 'the exams of the students' -is typical of RH 27 and is also found in pre-modern written Hebrew. 28 The 'analytic construct' is attested in early spoken Hebrew. 29 In Oded ha-Noded and Tsabar there are no instances of the 'analytic construct' but only of the synthetic construct, three of them in the former film and six in the latter. For example: The 'analytic construct' is another phenomenon that is attested in spoken Hebrew of the Mandate period but is not represented in the films of the time.
Independent object pronoun In contemporary Hebrew, the dependent object pronoun -for example, ‫היתבהא‬ as opposed to ‫יתבהא‬ ‫התוא‬ 'I loved her' -hardly ever occurs in the spoken register, and is becoming increasingly rare in the formal register as well. 30 In BH and RH, pronominal objects are generally suffixed to the verb, while independent pronouns appear only in marked contexts. 31 In Hasidic Hebrew, independent pronouns are slightly more frequent than dependent pronouns. 32 Evidence from the early 1930s indicates that, even then, the dependent object was rarely used in speech. 33 This fact can be given both a language-internal explanation and a language-external one. From an internal perspective, it can be argued that the analytic construction exists in the Hebrew sources and that the tendency to favour it over the synthetic construction is part of MH's general preference for analytic forms. The addition of dependent pronouns sometimes requires a modification of the verbal base, while the use of independent pronouns does not require this and is therefore simpler. The language-external explanation involves the influence of languages that were in contact with MH. 34 The films from the 1930s show a clear preference for the dependent object pronoun. In Oded ha-Noded seven of the eight pronominal objects are dependent pronouns, most of them appearing in the speech of the adult Jewish characters, such as: The film includes only one instance of the independent pronominal object, uttered by an adult Bedouin: 3. ‫ולא!‬ ‫חיש-מהר,‬ ‫אותם‬ ‫מצאי‬ ‫לכי‬ 'Go find them right away, or else!' The use of the independent pronoun in the speech of the Bedouin character might have been intentional, to signify 'otherness'. Tsabar includes one pronominal object, which is a dependent pronoun: We see that independent object pronouns, which evidently abounded in the Hebrew speech of the time, are not represented in the films produced in this period. In films from the 1950s independent object pronouns become dominant, but dependent ones continue to be used as well; in films from the 1960s, dependent pronouns disappear almost completely. 35 Independent possessive pronoun Spoken MH favours independent possessive pronouns over dependent ones (for instance, ‫רפסה‬ ‫ילש‬ 'my book' as opposed to ‫,)ירפס‬ although in certain circumstances dependent pronouns are the default. 36 In Biblical Hebrew (BH), only dependent possessive pronouns are attested. They continue to be dominant in RH, although independent possessive pronouns exist as well. 37 In Hasidic Hebrew dependent and independent pronouns are used with similar frequency. 38 The preference for the independent pronouns in the spoken language dates back at least to the Mandate period; there is evidence that the dependent possessive pronouns were rarely used in the speech of the time. 39 There were even those who proposed to abolish the dependent forms altogether and leave only the In Oded ha-Noded the independent pronoun is uttered by a child and is used to expresses contempt. This conforms to Ornan's observation that independent pronouns are obligatory in certain marked contexts, one of them being ironic speech. 42 This line in the movie is part of the cliquish humour of the native Eretz Israelis at the expense of non-natives 43 and is meant to favourably contrast the native boy with the ridiculous figure of the tourist. 44 Both films, then, display only a handful of independent possessive pronouns, although these were common in the speech of the Mandate period. Instead, they use the dependent forms that were dominant in written Hebrew. In later films, independent possessive pronouns become increasingly common at the expense of the dependent ones, although the latter continued to feature in the dialogue of Israeli films as late as the 1980s. 45

Polar question without interrogative particle
In contemporary spoken Hebrew, polar questions are usually formed without an overt interrogative particle and are similar in their syntactic structure to an indicative sentence. Only the rising final intonation marks them as having interrogative force; for example, ‫בירושלים?‬ ‫גרה‬ ‫את‬ '[Do] you live in Jerusalem?'. 46 The Hebrew sources feature polar questions of this sort alongside questions introduced by an interrogative particle. 47 In BH the interrogative element ha-generally introduces polar questions. The interrogative particle ha'im appears in the Bible only twice, and its function as a simple interrogative marker is doubtful. 48 In RH ha'im is absent 49 while the interrogative ha-is rare; 50 the particle weḵi introduces positive questions that expect a negative answer, and klum introduces negative questions that expect a positive answer. 51 The particle ha'im became established as a polar marker in Medieval Hebrew and is also common in Hasidic tales. 52 It seems that during the Mandate period polar questions were frequently formed without an interrogative particle, as evident from the following passage in Avineri (1930) Oded ha-Noded has three polar questions, all of them introduced by an interrogative element: two with ha-and one with weḵi. In Tsabar only three out of nine polar questions are introduced by an interrogative element, two by weḵi and one by ha'im: The occurrence of the bare polar questions in Tsabar but not in Oded ha-Noded may stem from the fact that the latter is a talking movie, which can use intonation to convey interrogative force, whereas the silent film Oded ha-Noded cannot. In this domain, then, there is a slight difference between the two movies from the 1930s. Both movies feature polar questions introduced by an interrogative particle, which are typical of written Hebrew. However, the silent film Oded ha-Noded includes no bare polar questions, although these were common in the speech of the time, whereas the talking film Tsabar includes six such questions. In films from the 1950s the proportion of bare polar questions rises, and in films from the 1970s and 1980s they appear almost exclusively. 54

Future-form verb expressing imperative force
In contemporary spoken Hebrew, the future form of the verb (yiqtol) is often used instead of the imperative form (qtol); for example, ‫תשמעו‬ 'listen!' lit. 'you will listen' instead of ‫מעו‬ ִ ‫.ש‬ 55 This feature has its roots in pre-modern

Summary of Part I
Part I presented eight linguistic phenomena that, according to observations by linguists and prescriptivist grammarians, were typical of spoken Hebrew of the Mandate period, and had roots in pre-modern written texts. 60 These phenomena are absent or only scantly represented in the films of the 1930s. Comparing the silent film Oded ha-Noded and the talking film Tsabar, one might expect the latter to include more instances of these colloquial phenomena, but this is not the case. Four of these phenomena -participles negated by lo, 'extra-positional' pronouns, definite-head constructs and 'analytic constructs' -are completely absent from both films. Two phenomena are found only in Tsabar: bare polar questions and the use of future forms to express imperative force. One phenomenon -independent object pronouns -is found only in Oded ha-Noded and another -independent possessive pronouns -is found in both films. A comparison of the two films shows that, even though Oded ha-Noded is a silent film, in which dialogue is presented using written captions, whereas Tsabar is a talking (albeit dubbed) film, the latter does not present a significantly higher incidence of the colloquial phenomena. 61 The language of the films from the 1930s, then, closely corresponds to the standard written language. Six of these phenomena were also examined in films from the 1950s and later, which were found to display a significant increase in their occurrence. Three phenomena become dominant to the complete exclusion of the alternative, namely: participles negated with lo, independent object pronouns and bare polar questions. Three other phenomena -definite-head constructs, independent possessive pronouns and future verbs with imperative forcebecome far more common than in the films from the 1930s, although the alternatives, typical of written Hebrew, are still found as well.

Part II: Phenomena in the process of exclusion from films' dialogue from the 1950s onward
This part of the article examines phenomena that are typical of contemporary written (rather than spoken) Hebrew. The prevalence of these phenomena in spoken Hebrew of the Mandate period is difficult to assess due to lack of evidence. Prescriptivist grammarians did not relate to these phenomena since both the more formal and the less formal alternatives are 'correct'. The formal phenomena gradually vanished from the dialogue of films produced in the 1950s-1980s. 62 The vanishing of these phenomena could possibly indicate that they belonged to the formal register even during the Mandate period.

Yiqtol form expressing present time
Yiqtol forms expressing present time are found in BH. 63 In MH such forms are encountered mostly in literary language. 64 In speech they are confined to certain fixed formal expressions, such as ‫לשתות?‬ ‫תרצה‬ ‫מה‬ 'Would you like to drink?' (lit. 'what will you want to drink'), ‫לי?‬ ‫לעזור‬ ‫תוכל‬ 'Could you help me?' (lit. 'will you be able to help me?'). 65 In Oded ha-Noded and Tsabar there are seven instances of yiqtol questions referring to the present, one in the former film and six in the latter. exhibit only a handful of yiqtol forms referring to the present, most of them in questions, and films from the 1970s and 1980s contain no such forms at all. 66 This formal phenomenon, then, is attested in films from the 1930s but gradually disappears in later films.

Predicate-subject word order
The unmarked word order in BH is predicate-subject, although subject-predicate order is also encountered in many constructions. MH (both spoken and written) retains the Biblical predicate-subject word order as the unmarked option in certain cases; for example, in impersonal constructions ‫לעשן(‬ ‫אסור‬ 'no smoking', lit. 'forbidden to smoke') and sentences of existence and possession ‫בחצר(‬ ‫כלב‬ ‫יש‬ 'there-is a dog in the yard'). Sentences introducing a new referent into the domain of discourse also tend to exhibit predicate-subject word order. 67 In written MH, sentences beginning with a fronted object or with an adverbial often display predicate-subject word order as well. 68 According to the formal rules of standard Hebrew, the predicate should be fronted only when it is a verb in the past or future tense, but not when it is in the present tense (i.e. a participle), 69 although the rules taught to Israeli schoolchildren often omit this nuance. 70 However, the unmarked word order in most types of MH sentence is subject-predicate, presumably due to the influence of European languages. 71 In Oded ha-Noded and Tsabar, most of the past-tense and future-tense sentences feature subjects that are incorporated in the verb. However, in both films, when the subject is independent, such sentences usually display a subject-predicate word order: This resembles the situation in RH, which likewise favours predicate-subject order in sentences where the predicate is a participle and the subject a pronoun in the first or second person. 72 To conclude, the films from the 1930s display many cases of predicatesubject word order, especially in nominal sentences and in present-tense verbal sentences. In the films from the 1950s, predicate-subject order is much less common. 73 Noun + bare proximal demonstrative MH uses the proximal demonstrative as a modifier in two main constructions, both of which place the demonstrative after the noun. The constructions are similar in meaning and are both semantically definite, but in one of them the noun and the demonstrative are both formally definite (i.e. marked with the definite prefix: ‫הזה‬ ‫האיש‬ 'this man', lit. 'the-man the-this'), while in the other neither the noun nor the demonstrative features a definite prefix: ‫איש‬ ‫זה‬ 'this man', lit. 'man this').
BH uses the formally definite construction, 74 whereas the construction with the bare proximal demonstrative occurs in the Bible only once. 75 Conversely, in RH the latter construction is dominant. 76 In Hasidic Hebrew the definite construction is used more frequently. 77 In MH both constructions are used, but the one with the bare proximal demonstrative is largely confined to writing. 78 The films from the 1930s featured both constructions, one in each film:  (Tsabar) In films from the 1950s and early 1960s the construction with the definite prefix is dominant, although the alternative construction, with a bare noun and demonstrative, still appears. In later films the latter construction disappears almost completely. 79 We again see that the films from the 1930s feature the construction that is typical of written Hebrew and that gradually fades from later films.

Participle suffixed with first-person pronoun referring to the subject
In MH several participles suffixed with a first-person pronoun are in use: for example, ‫תמהני‬ 'I wonder', ‫חוששני‬ 'I fear', ‫זוכרני‬ 'I remember'. This pattern is originated in RH. 80 In MH these forms are typical of formal and even archaic register, and limited to cognition verbs. 81 According to the Even-Shoshan dictionary, MH coined additional forms of this sort, such as ‫דומני‬ and ‫כמדומני‬ 'I think, I believe'. The films from the 1930s feature only one instance of such a form: 1. ‫שאלינו.‬ ‫דומני‬ ‫השיירה?‬ ‫מהי‬ ‫הבט.‬ ‫הנה,‬ ‫בוא‬ 'Come here, look. What is that caravan? I-think it's coming towards us.' (Tsabar) The entire corpus of films from the 1950s onwards yields only two more participles of this sort. 82 We see represented in the films from the 1930s, then, another phenomenon which is confined to formal register today and gradually vanished from the dialogue of later films.

The subordinator 'ašer
In BH the dominant subordinator is 'ašer, whereas in RH it is superseded by še-, which appears in BH only rarely. 83 MH has three main subordinators: 'ašer, še-and ha-; the first two appear in all syntactic environments, whereas ha-is confined to positive subject relative clauses in which the predicate is a participle. The subordinator še-is dominant to the spoken language, whereas 'ašer and ha-are largely confined to the written language. Even in writing the frequency of 'ašer has declined significantly since the revival period. Our 1930s' films include one relative clause featuring 'ašer and one featuring še-, neither of them preceding a participle: completely. As for free relative clauses and temporal clauses with 'ašer, they do not occur at all in the dialogue of films from the 1950s and onwards, although other types of narrative in the films, such as speeches, do contain a handful of examples. 85 We see, then, that the relativizer 'ašer, which is typical of written contemporary Hebrew, appears in films from the 1930s but disappears from the dialogue of later films.

The subordinator ki in content clauses
In written contemporary Hebrew, complement clauses are often introduced by the subordinator ki, which originates in BH, whereas the spoken language prefers the complementizer še-, which originates in RH. 86 The two films from the 1930s each feature one instance of a complement clause beginning with ki: from the 1960s and onwards exhibit none at all. 87 We see once again that the films from the 1930s display a phenomenon typical of written contemporary Hebrew, which disappears in later films.

'al + infinitive to express personal obligation
Another construction that is typical of the formal register in MH is 'al lit. 'on' + subject + infinitive to express personal necessity/obligation; for example, ‫ללכת‬ ‫עלי‬ 'I must go', lit. 'on me to go'. In spoken MH, personal obligation/necessity is generally expressed using the construction subject + ṣariḵ 'have to' + infinitive; for example, ‫ללכת‬ ‫צריך‬ ‫אני‬ 'I have to go'. The pattern with 'al is found in the Bible as well as in RH. 88 The patterns with ṣariḵ are first attested in RH. 89 The films from the 1930s use only the 'al + infinitive construction, which appears twice, both times in Oded ha-Noded: Later films mostly use the subject + ṣariḵ + infinitive construction to express necessity/obligation; the 'al + infinitive construction still makes an occasional appearance in films from the 1950s and early 1960s, but in later films it is absent. 90 The films from the 1930s again display a construction typical of written Hebrew that is not used in films made in later periods. I used to visit the synagogue'. 91 BH has two infinitive forms, the infinitive absolute (e.g. ‫מֹור‬ ָ ‫)ש‬ and the infinitive construct (e.g. ‫ֹר‬ ‫מ‬ ‫.)שְ‬ The infinitive absolute does not take prefixes or suffixes, whereas the infinitive construct can appear either in its bare form or prefixed with a preposition (e.g. ‫שמור,‬ ‫כשמור‬ ‫בשמור,‬ ‫.)משמור,‬ It can also be inflected ‫לשמרו(‬ 'to keep it') and appear as the head of a construct ‫אּול(‬ ָ ‫ׁש‬ ‫אֹות‬ ְ ‫ר‬ ‫כִ‬ ‫,וְ‬ 'and when Saul saw', lit. 'and at Saul's seeing', 1 Samuel 17:54) or as the annex in a construct ‫ָך(‬ ‫לְ‬ ‫כָ‬ ‫אֲ‬ ‫יֹום‬ ‫ּבְ‬ 'the day you eat', lit. 'in the day of your eating', Genesis 2:17). The infinitive construct frequently follows the prepositions be-'in' or ke-'as' as a minimal clause, usually a temporal one. RH retained only the infinitive construct prefixed with le-. 92 In MH, infinitive constructs without le-generally introduce temporal adverbial clauses. The infinitive construct is always prefixed with a preposition and followed by the performer of the action, represented by a noun or a dependent pronoun. 93 The films from the 1930s include one instance of an infinitive construct suffixed with a dependent pronoun without le-: The films from the 1950s yielded a handful of infinitive constructs without le-: a single instance in dialogue and several more in non-dialogue (for example, in a formal school play). 94 Here too we find in the films from the 1930s a phenomenon which disappears from later cinematic dialogues and in contemporary Hebrew is limited to formal register.

Summary of Part II
In the second part of this article I presented linguistic phenomena that are rarely found in contemporary spoken Hebrew, but whose prevalence in spoken Hebrew of the Mandate period is difficult to assess due to lack of evidence. Two of these phenomena were represented in Oded ha-Noded but not in Tsabar: 'al + infinitive to convey necessity/obligation, and relative clauses introduced by 'ašer. Conversely, three of the phenomena appeared in Tsabar but not in Oded ha-Noded: noun + bare proximal demonstrative, participles suffixed with a subject pronoun, and inflected infinitives without le-. Three phenomena appeared in both films: content clauses introduced by ki, yiqtol forms denoting the present tense, and predicate-subject word order in present-tense sentences. The last two phenomena are more prevalent in Tsabar than in Oded ha-Noded.
All the phenomena discussed in this part also appear in films from the 1950s, but very infrequently, and they disappear completely or almost completely from later films. A comparison between the talking film Tsabar and the silent film Oded ha-Noded reveals no significant difference between them in terms of the prevalence of these phenomena that were in the process of gradual exclusion from the language of films.
There is no evidence that the phenomena discussed in this section did not belong to spoken Hebrew during the Mandate period. However, these phenomena are suspected as being highly registered features even then. There are a few reasons to assume this: (1) in the Mandate period stratification already existed and, in many ways, resembled contemporary stratification; 95 (2) these phenomena were ejected from later films; (3) in contemporary Hebrew these phenomena are limited to formal register.

Discussion and conclusions
This article has examined the dialogue in two films from the 1930s, to determine whether it reflects spoken Hebrew of the time. It also has compared the dialogue in the two films, which were the first Hebrew silent film and the first Hebrew talking film, both produced in 1932. In addition it has compared the dialogue in these films with the dialogue of later films in terms of the representation of phenomena that were in process of exclusion from spoken language.
The first part examined phenomena that, according to testimonies in grammar books and prescriptivist writings, were characteristic of colloquial 95. Reshef, Hebrew in the Mandate Period.
Hebrew during the Mandate period. It showed that these phenomena were not represented at all in the films from the 1930s, whereas in films from the 1950s and onwards there is a significant rise in their frequency. The second part examined phenomena that are not typical of spoken language today but are represented in the dialogue of the films from the 1930s. These are phenomena whose prevalence in spoken Hebrew of the Mandate period cannot be judged from testimonies from that time. All of them are also found in films from the 1950s, but their frequency decreases and in later films they become very rare or vanish completely.
A comparison between the silent film Oded ha-Noded and the talking film Tsabar, both from the 1930s, showed that, contra to expectation, the dialogue in the talking film does not seem to correspond more closely to spoken Hebrew of the time. The two films do not differ in their representation of phenomena known to be typical of spoken Hebrew at the time, or in their representation of phenomena typical of written Hebrew, which gradually disappeared from later films.
An analysis of the findings shows that the language used in the films from the 1930s is similar to the language used in other cultural works of the period; 96 it is high-register Hebrew that, rather than reflecting the spontaneous speech of the time, aspires to conform to linguistic patterns that were thought to be ideal. As this study indicates, Eretz-Israeli film-makers of the period did not regard it as the function of cinematic dialogue to authentically reflect spoken speech. Several reasons for this might be suggested.
First, it seems that cinema was perceived as a cultural product that -like literature and theatre -merits 'refined' rather than colloquial language. The fact that speaking films were a new invention at the time and their production still involved considerable technical difficulty only heightened their perception as productions requiring formal speech.
Second, Hebrew at the time was still new as a spoken language, so dialogue in Zionist films had the important pedagogical function of presenting the desirable forms of speech. It is a case of mutual feedback: the register underscores the importance of the message, and the importance of the message dictates the register. In other words, the use of formal language signalled to the audience the importance of the content, and the importance of the content motivated the use of formal speech.
Third, since the directors of the films Oded ha-Noded and Tsabar were not themselves native speakers of Hebrew, they may have been oblivious to the stilted character of the dialogue.
Finally, we may assume that even if the film-makers of the 1930s wished to represent the authentic spoken Hebrew of their day rather than the linguistic ideal -and thereby challenge prevailing perceptions regarding colloquial language -Hebrew cinema in the 1930s was not yet sufficiently established to be a vehicle of such a subversive approach. During this period, local cinema was subject to censorship and suffered from economic dependence that led to interference in the content of films, which may have indirectly affected the character of the dialogue, among other aspects. The mere production of a Hebrew film was a big innovation, and it is unlikely that film-makers had either the necessary awareness or the motivation to be innovative in the representation of speech, which is only a small part of cinema's expressive capacity.