REGULARIZATION IN THE CRYSTALLIZATION OF MODERN HEBREW: THE CASE OF COUNTERFACTUAL CONDITIONALS 1

A BSTRACT Regularization is a process of linguistic reduction through the elimination of variants. Regularization processes occur naturally during language acquisition and learning. In social situations where learners comprise a large portion of the language community, regularization can lead to linguistic change. This was the case during the development of Modern Hebrew. Therefore, regularization processes are essential to a fundamental question about the crystallization of Modern Hebrew: to what extent its grammar continues the grammar of the previous layers of Hebrew and to what extent it features novel characteristics of its own . This paper focuses on the crystallization of counterfactual conditionals in Modern Hebrew. It shows that this process involved no new linguistic phenomena but only a culling of the large inventory of variants. These variants that coexisted during the revival period were all inherited from the preceding stages of Hebrew. A regularization process, which occurred mainly in the Mandate period, eliminated some variants, such as the positive meaning of ʾilmale and the qatal (regular past-tense) form in the main clause (the consequence). The variants that survived the regularization process underwent differentiation, becoming associated with distinct registers or meanings.


INTRODUCTION 1.1 Regularization in the Crystallization of Modern Hebrew
Regularization processes involve the reduction of linguistic diversity through the elimination of variants. It occurs naturally in individuals' language in language acquisition and learning. Regularization processes happen in all languages and periods, but there are situations where regularization brings about linguistic changes. These changes can occur in social situations where learners comprise a large portion of the language community. 2 This was the case during the early development of Modern Hebrew (MH).
The Hebrew language is a unique case of a language that stopped being used as a vernacular and did not have native speakers for about 1700 years. During these years, Hebrew was used as a part of diglossia and served mainly as a written language. At the end of the 19th century, an effort began to turn Hebrew into a national language that Jews would use in all areas of life. At first, MH did not have native speakers, and the language community comprised only L2 speakers of Hebrew. 3 The unique social situation in which Hebrew was returned to use in all fields of life raises the question of how the linguistic character of MH was shaped .
A fundamental question concerning the crystallization of MH pertains to the development of its syntax, which is the least conscious part of the language: how much of this grammar is inherited from the previous layers of Hebrew, and how much of it is new. 4 This paper examines this question through the concept of linguistic change through regularization processes in a situation where most speakers are not native. This question is explored here through the lens of a specific case: the regularization of counterfactual conditionals in MH.
The term "regularization," as used in this paper, does not encompass all the changes that occurred in MH but only the rapid changes that took place in the first decades of its development and shaped its character as we know it today. This period encompasses the revival period, which began in the 1880s and lasted some 40 years, roughly until 1920, and the Mandate period, which ended with the founding of the state in 1948. The term "contemporary Hebrew" refers to the language of the more recent decades, from around 1950 until today. 5 The departure point for the regularization of various grammatical forms in MH was the considerable variability that characterized the language of the revival period. 6 This variability stemmed, inter alia, from the unique ways in which knowledge of Hebrew was transmitted throughout the ages, due to which all phases of the language were simultaneously accessible to its users, especially educated writers. 7 The mechanism presented here shows that the syntax formed in the first decades of MH is based on the syntax of Hebrew in the previous layers. MH's syntax was developed mainly not through innovations of syntactic structures but rather through a selection from existing options. The regularization process involved the decline or disappearance of certain linguistic characteristics among the inventory of existing phenomena, whether inherited from previous stages or borrowed from the substrate languages and the reinforcement of other characteristics. 8 In the early stages of MH, this culling process was relatively rapid and complete and took place mainly during the Mandate period. 9 Some of the variants that "survived" were replaced in later stages, but more slowly and less completely. Once the regularization process was over, any competing forms that remained underwent semantic or register differentiation, as two hegemonic types of Hebrew emerged: institutional (planned) Hebrew  Hebrew, 10 alongside other language genres and types. 11

The Case of Counterfactual Conditionals
In contemporary spoken Hebrew, the common and unmarked conditional marker in both factual and counterfactual conditionals is ʾim ("if"); in negative conditionals, it is accompanied by the negative element lo. In standard Hebrew, on the other hand, ʾim is confined to factual conditionals (whereas in counterfactual conditionals, it is considered incorrect). 12 Counterfactuals use designated conditional markers: lu and ʾilu appear in positive counterfactuals and are only rarely paired with the negative marker lo, 13 while negative counterfactuals feature the markers lule, ʾilmale, and ʾilule, the first two of which are fairly common and the third more rare. 14 These three negative counterfactual markers are now regarded as synonymous. 15 They appear in the same syntactic environments and have the same meaning, "if not," whether they occur before a noun phrase or before a conditional clause. All three are now spelled with the letters lamed-aleph at the end, identical to the spelling of the negative element lo, reflecting their negative meaning. However, in the revival period there were considerable differences between them, reflecting their usage in previous layers of Hebrew. 16 This paper will describe and analyze the use of counterfactual conditionals during the revival period and the regularization of their use in MH. Previous studies have examined counterfactual conditionals in the pre-modern stages of Hebrew. 17 Their use in MH has been mainly discussed from a prescriptive viewpoint, 18 alongside synchronic descriptions of their use in contemporary Hebrew. 19 However, no study has provided a synchronic description of their use in the revival period or a diachronic description of their regularization in MH. This paper will do so while showing the relation to overall regularization processes in MH. In a previous article, I examined negative counterfactual conditionals based on other corpora. 20 In this article, I will focus on the positive counterfactual conditionals. The previous article's conclusions will be incorporated here.
Data for this study was sourced from two main corpora. The description of counterfactual conditionals in the revival period is based on a corpus of texts dating from 1882 to 1914, available on the website of the Hebrew Academy's Historical Dictionary Project. 21 This corpus, comprising academic texts by seven different authors, yielded 148 positive counterfactual conditionals sentences. 22 Data for the period between the revival and contemporary periods, during which the regularization process took place, is sourced from the Haaretz newspaper corpus. It consists of issues of the Haaretz daily from 1920-1960. The issues, from the first three months of round years (1920, 1930, etc.), yielded 868 examples of counterfactual conditionals. 23 The description of the situation in contemporary Hebrew is based on previous studies, 24 as well as on hundreds of examples of counterfactual conditional sentences collected from literary, cinematic, and television texts and from the Internet, dating from 1960 until the present day. A systematic sample test was conducted in the Maariv newspaper from the first ten days of January 1980 and 1990.
The first part of the paper describes the use of counterfactual conditionals during the revival period and traces its sources in the preceding layers of Hebrew. The second part of the paper describes the regularization of their use in MH, analyzing the circumstances in which this process took place and the factors that affected it. The third part of the paper summarizes the characteristics of the regularization process and examines it as an example of the overall regularization process of grammatical constructions in MH.

COUNTERFACTUAL CONDITIONALS IN THE REVIVAL PERIOD
The revival period was characterized by considerable variability (compared to the more limited and uniform situation today), reflecting the breadth and diversity of the material inherited from the previous stages of Hebrew. The section below examines whether counterfactual conditionals during the revival period reflect their use in Classical or Medieval Hebrew and whether revival period literature reflects any new uses. 22. The academic language of the revival period is much closer to contemporary Hebrew than the literary language of that time. However, it too contains linguistic phenomena that were later discarded (Y. Reshef

Counterfactual Conditionals Markers in the Revival Period
During the Revival period, there was a variety of conditional markers in counterfactual conditionals. In positive sentences, the main use was of designated conditional markers: lu and ʾilu and also ʾilmale in a positive meaning ("if"), and the general conditional marker ʾim was rare. Three markers were used in negative sentences: lule, ʾilule, and ʾilmale in a negative meaning ("if not"). 25 In the revival period, lule, ʾilule and ʾilmale are always negative when preceding a noun phrase, but when preceding a clause, they differ in their meaning: lule and ʾilule are consistently negative, whereas ʾilmale can have a negative sense ("if not," as in Example 1) but also a positive one ("if" as in Example 2); the second of these options is the more common one. 26 1.
‫אלמלי‬ ‫זעומה‬ ‫שאגה‬ ‫שחצנית,‬ ‫נזיפה‬ ‫במקום‬ ‫יוצאת,‬ ‫שהיתה‬ ‫אפשר‬ ‫לנזוף,‬ ‫באים‬ ‫היינו‬ ‫אנו‬ ‫החשוכה‬ ‫הסביבה‬ ‫אותה‬ ‫של‬ ‫העורון‬ ‫על‬ !... "Had we come to rebuke, perhaps instead of an arrogant rebuke, a furious roar would have erupted over the blindness of the dark surroundings…!" 28 The two opposite meanings of ʾilmale often create ambiguity, but the context can help infer the meaning, and so can the spelling. During this period, the negative markers were spelled both with a final aleph ‫לולא(‬ ‫אלולא,‬ ‫)אלמלא,‬ and with a final yod ( ‫לולי‬ ‫אלולי,‬ ‫אלמלי‬ , ). The spelling of lule and ʾilule is unrelated to their polarity, for they are always negative, but the spelling of ʾilmale is closely correlated with its polarity. When spelled with a yod (as in example 2), it is nearly always positive, and when spelled with an aleph (as in example 1), it is nearly always negative. 29 25. lo comes after lu or 'ilu rarely. 26. Since ʾilmale in the revival period had a positive meaning alongside the negative one, in rare instances it was used as an optative marker in monoclausal constructions (without a consequent). The most common positive markers in the revival period were the designated counterfactual conditional markers: lu was the most common, and after it, ʾilu. The neutral conditional marker ʾim is the third most common in the corpus, but it occurs only in the writing of one writer -Ahad Ha'amout of the seven writers in the corpus. ʾilmale in its positive meaning occurs only in 7% of the cases (11 times), but it appears in texts of four different writers.
Chart 1: The positive counterfactual markers in the revival period Revival-period Hebrew inherited all counterfactual conditional markers from the preceding stages of the language: Biblical Hebrew uses the marker lu (and lule), and the markers ʾilu, 30 and ʾilmale (and ʾilule), emerged in Rabbinic Hebrew. In Biblical Hebrew, ʾim is used as a conditional marker only in factual and concessive conditionals. 31 In the Interim period between the Rabbinic period and the revival era, during which Hebrew was used solely as a literary language, the use of ʾim in counterfactual conditionals is common. 32 How did ʾilmale come to have two opposite meanings in the revival period? Apparently, all three markerslule, ʾilmale, and ʾilulewere originally negative. This assumption is supported by the fact that, according 30 to most etymological analyses, all three contain the negative element lo. 33 Furthermore, when preceding a noun phrase, all three are always negative, in all stages of Hebrew that feature them, and the same is true for lule and ʾilule followed by a clause. 34 The only exception is ʾilmale, which in Rabbinic -Babylonian Hebrew is nearly always positive when followed by a clause. 35 This development is generally explained in diachronic terms: the marker was originally negative in all syntactic environments, but a semantic shift produced a positive meaning in pre-clausal positions only. 36 In the Medieval period, various writers reverted to the (original) negative meaning of ʾilmale. 37 As a result, in the modern Hebrew literature of the pre-revival period, the negative sense was prevalent again, alongside the positive one. 38 The orthographic situation likewise has its roots in the previous layers of Hebrew. In Classical Hebrew, the spelling of the markers alternated between aleph and yod regardless of their meaning. 39 The spelling with aleph is 33. Lule is generally analyzed as the counterfactual conditional marker lu +the negative marker lo,and ʾilule as the conditional marker ʾi +lu+lo (M.Z. Segal, The spelling with yod results from a process of dissimilation between the vowel u in lu and the vowel o in lo, which transformed the last vowel into e. This dissimilation, in turn, triggered the change in spelling, causing the aleph to be replaced with yod, which is more usual following a consonant vocalized with tzere. 40 The correlation between positive ʾilmale and the spelling with yod was a conscious decision made in the Medieval period. Since the ambiguous polarity of ʾilmale posed a problem for readers of Hebrew, Rabbeinu Tam, of the 12 th century, suggested that positive ʾilmale is spelled with yod and negative ʾilmale with aleph. Editors and proofreaders later applied this convention to earlier texts, thus creating the impression that this correlation between the spelling and the meaning had always existed. 41

Counterfactual conditionals verbal forms in the revival period
Conditional sentences consist of a clause expressing the condition (antecedent), usually introduced by a conditional marker, and a clause expressing the consequence (consequent). Revival-period Hebrew had three main sequences of verbal forms in counterfactual conditional, all of them comprising a past-tense antecedent and a past-tense consequent: a. Haya-haya: Compound past, consisting of the verb haya (past tense of be) followed by a present participle form (qotel), in both the antecedent and the consequent 3. The counterfactuality can be expressed either by using the compound past in the antecedent (which allows using the unmarked conditional element ʾim), or alternatively by employing a designated counterfactual marker (which allows using the simple past in the antecedent), or by combining both. When the antecedent precedes the consequent, the latter nearly always begins with ki ʾaz or ki ʿata when followed by a qatal form (example 5) or, less frequently, by a compound past form or some other predicate (example 6). When the consequent precedes the antecedent, it is not introduced by a marker (example 7). As we will see below, specific markers appear with certain sequences of verbal forms . 6. ‫לו‬ ‫מורים‬ ‫היו‬ ‫האלה‬ ‫לאנשים‬ ‫כי‬ ‫ההרחבה‬ ‫והחדוש‬ ‫אמנם,‬ ‫טובים‬ ‫בשעה‬ ‫רק‬ ‫אך‬ ‫ש‬ ‫הם‬ ‫היתה‬ ‫אז‬ ‫כי‬ ‫להחרוזות,‬ ‫צריכים‬ ‫שהם‬ ‫בשעה‬ ‫ולא‬ ‫השפה‬ ‫לגופם,להתעשרות‬ ‫צריכים‬ ‫ועשירה‬ ‫יפה‬ ‫שפתנו‬ ‫הזה.‬ ‫כיום‬ Had these people been instructed that expansion and innovation are good indeed, but only when needed to enrich language and not for rhymes, then our language would be beautiful and rich today. 45 7.

‫היו‬ ‫אילו‬
‫האלה‬ ‫היתרונות‬ ‫ושני‬ ‫יכלו‬ ‫הרבה,‬ ‫שפתנו‬ ‫את‬ ‫להעשיר‬ ‫לו‬ ‫נמצאו‬ ‫כראוי‬ ‫בהם‬ ‫משתמשים‬ And these two benefits could have enriched our language a lot, had they been appropriately used. 46 In addition to these three main sequences of verbal forms, the markers also appeared in several other patterns in revival-period Hebrew. The three most common sequences of verbal forms typically relate to the past and express a counterfactual conditional that has not occurred in the past. In rare cases, they represent a hypothetical condition. In contrast, yiqtol-yiqtol represents only hypothetical conditionals.
Revival-period Hebrew inherited all counterfactual conditional markers from the preceding stages of the language, along with their usage: The Biblical markers appear with Biblical sequences of verbal forms and the Rabbinic markers with Rabbinic ones. Biblical Hebrew uses the marker lu with qatalqatal to expresses past counterfactual conditionals. The consequent can be 47 introduced by ki or by ki+ʾaz/ʿata, but these markers are non-obligatory. 49 The markers ʾilu and ʾilmale, and the haya-haya and qatal-haya sequences, emerged in Rabbinic Hebrew. 50 In Biblical and Rabbinic Hebrew, the markers also appear with other sequences of verbal forms, such as sentences expressing hypothetical future or present conditionals, employing qotel or yiqtol forms in both the antecedent and the consequent. 51 Revival-period Hebrew keeps the correlation between the marker and the sequence of verbal forms according to the original layer. Biblical sequences favor biblical markers, and Rabbinic sequences prefer Rabbinic markers. In this period, both lu and ʾilu appear with qatal-haya. They differ in their distribution with the other two patterns, however: ʾilu prefers haya-haya, whereas lu distinctly favors qatal-qatal. The negative markers lule and ʾilmale also preserve the correspondence between the origin of the marker and the forms' sequence: the biblical lule occurs more frequently with biblical qatalqatal, whereas the Rabbinic ʾilmale occurs more regularly with Rabbinic haya-haya. 52 In Biblical Hebrew, ʾim is used as a conditional marker only in factual and concessive conditionals. 53 ʾim is found in counterfactual conditionals in some Babylonian Talmud manuscripts, but it is difficult to know whether it is an original or late use. 54 In the interim period, during which Hebrew was used solely as a literary language, the use of ʾim in counterfactual conditionals is common. 55 In this period, the counterfactual conditional was generally expressed using past tense verbs, but sometimes also using other forms, mainly yiqtol. 56 In the modern writings that preceded the revival of spoken Hebrew, counterfactuals were very diverse. In the Hebrew of the Enlightenment period, the most common pattern is qatal-qatal, 57  In conclusion, during the revival period, the markers and sequences of verbal forms all originate in the previous layers of Hebrew. Furthermore, the correspondence between the origin of the marker and the sequence of verbal forms is preserved. In other words, revival-period Hebrew adheres to the original constructions found in previous texts without breaking them into independent components.
The most common patterns in this period are the Rabbinic haya-haya and qatal-haya, both of which have compound past forms in the consequent, and the Biblical qatal-qatal, which has a simple past-tense form in the consequent, but which, unlike in Biblical Hebrew, must include ki in the consequent. Some of the counterfactual conditionals' characteristics in the revival period continue Medieval Hebrew use, including the use of ʾim and the distinction between the two meanings of ʾilmale through spelling. 59

THE REGULARIZATION OF COUNTERFACTUALS CONDITIONALS IN
THE POST-REVIVAL PERIOD As seen, Revival-period counterfactuals conditionals were characterized by linguistic diversity that was inherited from previous layers of the language. During the Mandate period, regularization processes took place, reducing this variety by eliminating variants of both markers and patterns.

The Regularization of the Counterfactual Conditional Markers
As stated, the four positive counterfactual conditionals markers in the revival period were lu, ʾilu, ʾim, and positive ʾilmale. In 1950, after the Mandate period, the two common markers were lu and ʾilu (ʾilu was much more common). The other two markers -ʾim and positive ʾilmalewere extremely rare in the corpus. But the explanation for the rarity of each is different .
The disappearance of positive ʾilmale largely took place during the Mandate period. 60 At the time of the state's founding, clause-initial ʾilmale was predominantly negative, although occasional instances of the positive marker still occurred. Sporadic instances of positive clause-initial ʾilmale persisted even in later periods, in texts by revival-generation writers. 61 The pace at which regularization occurred differed from genre to genre, and there are even differences within genres, for instance, between newspapers aimed at different sectors. Despite the differences between genres, and the residual late occurrences, we can say that positive ʾilmale has effectively disappeared from the language. The fact that contemporary native speakers unschooled in the relevant Hebrew literature are unfamiliar with positive ʾilmale indicates that it has virtually become extinct. 62 Why was the positive meaning of ʾilmale discarded? As mentioned, the existence of two opposite meanings was a source of confusion already in the Medieval period. It led to the introduction of a rule for differentiating between them in writing. However, with the revival of spoken Hebrew, the confusion emerged again since the two meanings are pronounced identically. 63 The potential for confusion led the language authorities to recommend using ʾilmale in negative contexts only. 64 This case is an unusual one, in which the authorities rejected a literary form that has its roots in Classical sources. The language authorities' preference for negative ʾilmale and their rejection of the positive use, prevalent in the Babylonian Talmud, may have been influenced by pre-revival Hebrew literature, which used positive ʾilmale but distinctly favored the negative one. 65 Since ʾilmale was mostly confined to the written language, the opinion of the language authorities apparently carried some weight. The survival of the negative meaning in MH seems to have been motivated by two additional factors: the etymology (the presence of the element lo) and by analogy to lule and ʾilule. The impact of these factors was already evident in written Hebrew of the pre-revival period and grew stronger after the revival of Hebrew as a spoken tongue.
The decline of the positive meaning was accompanied by a decrease of the yod-final spelling, and by a growing dominance of the etymological spelling with aleph, not only in the case of ʾilmale but also in the case of lule and ʾilule. This process too occurred in the Mandate period. During the revival period and the Mandate period, Rabbeinu Tam's rule, (i.e., aleph-final negative, yodfinal Positive) was followed very strictly. But in the post-1948 period, this situation changed drastically. Aleph-final ʾilmale continued to denote the negative meaning, but yod-final ʾilmale was more frequently negative as well. In other words, while in the revival period, the spelling with yod was a fairly reliable indication of positive meaning, it lost this function in later periods. 66 Unlike positive ʾilmale, it seems that ʾim was common in the spoken language in counterfactual conditionals early in MH. Its rarity in the written corpus in 1950 reflects a prescriptive approach that opposes its non-classic use in counterfactual conditionals. Mordechai Ben-Asher wrote in 1972 that all the grammar books he examined do not mention the use of ʾim in counterfactual conditionals, and thus they ignore the common use in Modern Hebrew. 67  Despite the opposition, the use of ʾim in counterfactual conditionals became very dominant in contemporary Hebrew. In the Maariv corpus in 1980, ʾim occurs in 11% of the counterfactual conditionals (8 times), while in 1990, it appears in 30% of the cases (25 times). This process reflects the weakening of the prescriptive approach, which allows colloquial-language characteristics in artistic and written language. 69 In sum, looking at a written corpus from the Mandate period, it appears that ʾim and positive ʾilmale were ceased to be used, leaving only two positive counterfactual conditionals markerslu and ʾilu. But from the perspective of our time, these findings seem puzzling. Positive ʾilmale is not used today at all, whereas ʾim is very common in counterfactual conditionals. This case highlights the need to be extra careful in research that relies on corpora from the first decades of MH because of the dominance of the prescriptivist approach in this period, which can lead us to erroneous conclusions. 70 Therefore, external evidence from that period should be sought, as well as later corpora, which can help to shed light on findings from earlier corpora.

The Regularization of the Counterfactual Conditional Verbal Forms
As stated above, in the revival period, there were three main sequences of verbal forms to express counterfactual conditionals. The qatal-haya sequence, featuring a simple past form in the antecedent and a compound past form in the consequent, favors all conditional markers. The haya-haya sequence originates in the Rabbinic literature and favors the Rabbinic markerʾilu, and also ʾim; and qatal-qatal, which originates in the Bible appears most often with the Biblical lu. In the course of the regularization process, qatal-qatal was discarded altogether, leaving two options. 1950 in 12 constructions (5%) and in 1960 in only 3 constructions (1%). In contemporary Hebrew, when the antecedent precedes the consequent, the latter usually lacks any initial marker, though in rare cases it is introduced by the marker ʾaz. 76 Chart 5: The decline of ki (ʾaz/ʿata) in the consequent of counterfactual conditionals As stated, in the revival period, ki was mandatory in qatal-qatal, unlike in Biblical Hebrew, where it is optional, and despite the tendency in the revival period to emulate the language of the Bible. In later stages of Hebrew, ki lost its modal function and therefore ceased to be used as a means of modalizing the consequent in conditionals. This left only the two other means: the compound past or the modal verb yaxol. As a result, the qatal-qatal sequence, in which the consequent features a simple past verb, was ruled out, except when that verb is yaxol. 77 The yiqtol-yiqtol sequence, which was rare in the revival period, was also ruled out and practically not used in the Mandate period.
During the revival period, conditional markers appeared in the sequences of verbal forms in which they were used in their original layer of the language. This correlation was preserved in 1920 but later blurred. 78 At the end of the Mandate period, lu and ʾilu appear in the two remained sequences, qatal-haya, and haya-haya, without a significant difference between them. The three 76. T. Bar, "Conditional Clause," p. 550; M. Ben-Asher, Syntax of Modern Hebrew, p. 127. 77. L.H. Glinert, "The Hypothetical Conditional," p. 51. 78. In 1920 qatal forms occurred in the consequent of counterfactual conditionals 9 times with lu (15%) and not once with ʾilu (0%). In 1940, qatal forms occurred in the consequent once with lu (6%), and 7 times with ʾilu (5%)

DISCUSSION AND CONCLUSION
This paper describes the formation of grammatical construction in MH through regularization, i.e., a process that did not involve the creation of new linguistic phenomena but only the selection of certain options from a range of revival-era variants. The variants that remained after this process underwent differentiation in terms of their register and meaning.
Chart 8: Regularization in the Crystallization of Modern Hebrew During the revival period, counterfactual conditionals were characterized by variability in markers and constructions compared to today. The numerous variants that coexisted during the revival period originated in earlier stages of Hebrew. Biblical markers appear in the Biblical constructions, whereas Rabbinic markers appear in the Rabbinic constructions. This fact reflects adherence to the Classical sources and perhaps even dependence on them. Other revival-period characteristics reflect Medieval practices, such as the non-classic use of ʾim in counterfactual conditionals and the use of spelling to distinguish between the two meanings of ʾilmale, which is a Medieval convention but not a Classical one.
The regularization process involved discarding some variants: positive ʾilmale and the yod-final spelling, the qatal-qatal, and the yiqtol-yiqtol sequences of verbal forms. This has left us with two constructions, both with a compound past form in the consequent but differing in the antecedent structure: one has a compound past in the antecedent (haya-haya) and the other a simple past (qatal-haya). These two constructions have undergone

The Revival Period
Variability in markers and constructions; Affinity between marker and construction according to the original layer.

The Mandate Period
A regularization process involved discarding some variants; Less Affinity between marker and construction according to the original layer.

Contemporary Hebrew
Less variants; differentiation in register and meaning. differentiation in their register and meaning; haya -haya is less formal and can express hypothetical rather than just counterfactual conditional.
The regularization of the counterfactual conditionals largely occurred during the Mandate period, when Hebrew began serving as the main language of everyday life in the Yishuv. 80 However, some of the phenomena discarded continued to occur in later periods, albeit sporadically, such as positive ʾilmale and simple past in the consequent. This reflects the polychronic character of written MH. Written texts from the revival period, for example, often included archaic features that were not characteristic of the Hebrew of the time. Similarly, texts written in the first decades after independence could include revival-era features that had already been discarded from the language but persisted sporadically, especially in the language of older writers.