►II.21.A.i. THE LARYNGEALS ʔ, h
§α. The alternation between ʔ and h is not rare in Arabic and Semitic. For example, the verb ʔarȃda أَرَاْدَ “(he) wanted” has a variant harȃda هَرَاْدَ “(he) wanted” (the second variant is thought to be older, I will talk about this later). The word ʔin إِنْ “if” had a variant hin هِنْ in the dialect of Ṭayyiʔ طَيِّئٌ :
وَحَكَى اِبْنُ جِنِّيٍّ عَنْ قُطْرُبَ أَنَّ طَيِّئًا تَقُولُ “هِنْ فَعَلْتَ فَعَلْتُ”، يُرِيدُونَ “إِنْ”، فَيُبْدِلُونَ. [تاج العروس، “أن”]
Some Arabs (perhaps also the Ṭayyiʔ ?) said la-hinna-ka لَهِنَّكَ for la-ʔinna-ka لَئِنَّكَ “verily you(masc. sing.)” and hiyyȃ-ka هِيَّاْكَ for ʔiyyȃ-ka إِيَّاْكَ “you(object masc. sing.).”
قال الجوهري: وقولهم “لَهِنَّكَ”، بفتح اللام وكسر الهاء، فكلمة تستعمل عند التوكيد، وأَصله “لإنَّكَ” فأُبدلت الهمزة هاءً، كما قالوا في “إِيَّاكَ” “هِيّاكَ”، وإنما جاز أَن يجمع بين اللام وإنَّ وكلاهما للتوكيد لأَنه لما أُبدلت الهمزة هاء زال لفظ إنّ فصار كأَنه شيء آخر. قال الشاعر “لَهِنَّكِ من عَبْسِيَّةٍ لَوَسِيمَةٌ … على كاذبٍ من وعْدِها ضَوْءُ صادقِ”. اللام الأُولى للتوكيد والثانية لام إن. وأَنشد الكسائي “وبي من تَباريحِ الصَّبابة لَوْعةٌ … قَتِيلةُ أَشواقي وشَوْقي قَتيلُها … لَهِنِّكِ من عَبْسيَّةٍ لَوَسيمةٌ … على هَنَواتٍ، كاذبٍ مَنْ يَقُولُها” […] ومنه قول محمد بن مَسلمة “أَلا ياسَنا بَرْقٍ على قُلَلِ الحِمَى … لَهِنَّك من بَرْقٍ عَلَيَّ كَريمُ … لمَعْتَ اقْتِذاءَ الطيرِ، والقوْمُ هُجَّعٌ … فهَيَّجْتَ أَسْقاماً وأَنتَ سَلِيمُ”. واقْتِذاءُ الطائرِ هو أَن يفتح عينيه ثم يُغْمِضَهما إغْماضَةً. [لسان العرب “لهن”]
In Nuṣayri dialects of western Syria the pronouns ʔanȃ أَنَاْ “I,” ʔanta أَنْتَ “you(masc. sing.),” and ʔanti أَنْتِ “you(fem. sing.)” are pronounced hana, hint, and hinte (c.f. the Šeħri equivalents he, hɛt, and hit). The relative pronoun ʔalli/ʔilli has a variant halli in some Syrian dialects. The demonstratives hā-ðiy هَذِيْ “this,” hȃ-huwa هَاْ هُوَ “here he [is],” and hȃ-hiya هَاْ هِيَ “here she [is]” are pronounced in modern Egyptian Arabic ʔādi, ʔahoh, and ʔaheh.
§β. The sounds ʔ and h may be strengthened to the pharyngeal ħ and ʕ or vice versa. For example, Classical hȃ-huwa هَاْ هُوَ “here he [is]” and hȃ-hiya هَاْ هِيَ “here she [is]” are pronounced in Lebanon ʔaħħuwwe and ʔaħħiyye. The name of the Ancient North Arabian god Nhy was written ᵈNu-ḫa-a-a /Nuχaay/ in the “Thompson prism” (Nineveh A) of Esarhaddon, and Nħy ܢܚܝ in Syriac, so it appears that this name was pronounced *Nuħay. According to ʔAs-Suyȗṭiyy السيوطي (in المزهر في علوم اللغة وأنواعها ), some Arabians said ʕanna-ka عَنَّكَ for ʔanna-ka أَنَّكَ “that you(masc. sing.),” ʕAslamu عَسْلَمُ for the personal name ʔAslamu أَسْلَمُ , and ʕuðunuṋ عُذُنٌ for ʔuðunuṋ أُذُنٌ “ear.”
وَمِنْ ذَلِكَ الْعَنْعَنَةُ، وَهِيَ فِي كَثِيرٍ مِنْ الْعَرَبِ، فِي لُغَةِ قَيْسٍ وَتَمِيمَ، تَجْعَلُ الهَمْزَةَ الْمَبْدُوءَ بِهَا عَيْنًا، فَيَقُولُونَ فِي “أَنَّكَ” “عَنَّكَ”، وَفِي “أَسْلَمَ” “عَسْلَمَ”، وَفِي “أُذُنٍ” “عُذُنٍ”. [السيوطي، المزهر في علوم اللغة وأنواعها]
The Andalusian ʔIbn Sȋdah ابن سيده mentioned in المخصص numerous examples of variation between laryngeals and pharyngeals, e.g. the words ʔan أَنْ “that” and la-ʕalla لَعَلَّ “hopefully” had dialectal variants ʕan عَنْ and la-ʔalla لَأَلَّ .
وَيُقَالُ “أَرَدْتَ أَنْ تَفْعَلَ كَذَا وَكَذَا”، وَبَعْضُ الْعَرَبِ يَقُولُ “أَرَدْتَ عَنْ تَفْعَلَ”. وَقَالَ اِبْنُ الْسِّكِّيتِ: “لَأَلَّنِي” يُرِيدُ “لَعَلَّنِي”. [ابن سيده، المخصص، بَاب مَا يَجِيء مقولًا بحرفين وَلَيْسَ بَدَلًا]
The word ħatta·y حَتَّىْ “until” was pronounced ʕatta·y عَتَّىْ in the dialect of Ṭȃʔif and the nearby Huðayl هُذَيْلُ , and this was further weakened to ʔatta·y أَتَّىْ .
عُمَرُ رَضِيَ اللهُ تَعَالَى عَنْهُ قَالَ لِعَبْدِ اللهِ بْنِ مَسْعُودٍ حِينَ بَلَغَهُ أَنَّهُ يُقْرِئُ الْنَّاسَ “عَتَّى حِينٍ” [يُرِيدُ حَتَّى حِينٍ]: إِنَّ الْقُرْآنَ لَمْ يَنْزِلْ بِلُغَةِ هُذَيْل فَأَقْرِئْ الْنَّاسَ بِلُغَةِ قُرَيْش. [قَالَ] الْفَرَّاءُ: “حَتَّى” لُغَةُ قُرَيْش وَجَمِيعِ الْعَرَبِ إِلَّا هُذَيْلًا وَثَقِيفًا فَإِنَّهُمْ يَقُولُونَ “عَتَّى”. قَالَ: وَأَنْشَدَنِي بَعْضُ أَهْلِ الْيَمَامَةِ “لَا أَضَعُ الْدَّلْوَ وَلَا أُصَلِّي … عَتَّى أَرَى جِلَّتَهَا تُوَلِّي … صَوَادِرَ مِثْلَ قِبَابِ التَّلِّ”. وَقَالَ أَبُو عُبَيْدَة: مِنْ الْعَرَبِ مَنْ يَقُولُ “أَقِمْ عَنِّي عَتَّى آتِيكَ وَأَتَّى آتِيكَ” بِمَعْنى حَتَّى آتِيكَ، وَهِيَ لُغَةُ هُذَيْل. [الزمخشري، الفائق في غريب الحديث “عتى”]
See also Chaim Rabin, p. 201, §q.
§γ. According to the old linguists, the hamza·t did not exist in west Arabian dialects of Classical Arabic. The elision of hamza·t in those dialects was called tashȋlu ʔ·al-hamza·ti تَسْهِيْلُ اَلْهَمْزَةِ “simplification of hamza·t” or “alleviation of hamza·t.”
The orthography of the Qurʔȃn (before the dots and diacritics) reflects the Classical dialect of Mecca c. 650 AD. This orthography confirms that the hamza·t was dropped in the dialect of Mecca, because if we remove the hamza·t diacritic ء the word biʔ.ruṋ بِئْرٌ “well” will turn out to be bȋ.ruṋ بير , with the change iʔ > ȋ. Similarly, the word muʔ.mi.nuṋ مُؤْمِنٌ “believer” must have been pronounced mȗ.mi.nuṋ مومن , with the change uʔ > ȗ. Words such as raʔ.suṋ رَأْسٌ “head” and baʔ.suṋ بَأْسٌ “might” were probably pronounced rȃsuṋ راس and bȃsuṋ باس .
§δ. The word yas.ʔa.mu يَسْأَمُ “(he) becomes/will become bored” is written in the Qurʔȃn يسم , which reflects the pronunciation ya.sȃ.mu > ya.sa.mu (note that the elision of the syllable-initial hamza·t changed syllable boundaries). The word maš.ʔa.ma.·tuṋ مَشْأَمَةٌ “left” is written مشمة , which indicates the pronunciation ma.šȃ.ma.·tuṋ > ma.ša.ma.·tuṋ. The word yan.ʔaw.na يَنْأَوْنَ “they(masc. plur.) stay away” is written ينون , for the pronunciation ya.naw.na. The word ʔaf.ʔi.da.·tuṋ أَفْئِدَةٌ “hearts” is written أفدة , for the pronunciation ʔa.fȋ.da.·tuṋ > ʔa.fi.da·tuṋ.
§ε. The elision of a hamza·t at the beginning of a syllable that immediately followed a definite article ʔ·al– would have eliminated the need for the prosthetic ʔ·a– of the definite article. For this reason the word ʔ·al–ʔayka·ti اَلْأَيْكَةِ is sometimes spelled in the Qurʔȃn ليكة , for the pronunciation layka·ti.
§ζ. The phrase ʔa–ʔan.tum أَأَنْتُمْ “is it you(masc. plur.)?” is written انتم , for ʔȃn.tum. The word ya.ta.sȃ.ʔa.lȗ.na يَتَسَاْءَلُوْنَ “they wonder” is written يتسالون , for ya.ta.sȃ.lȗ.na. The word mun.ša.ʔȃ.tuṋ مُنْشَآتٌ “lofty(fem. plur.)” is written منشات , for mun.šȃ.tuṋ. The word bȃ.ri.ʔi.kum بَاْرِئِكُمْ “your(masc. plur.) creator(gen. masc. sing.)” is written باريكم , for bȃ.rȋ.kum. The word xȃ.ṭi.ʔȋ.na خَاْطِئِيْنَ “sinners” is written خطين , for xȃ.ṭȋ.na. The word ru.ʔȗ.suṋ رُؤُوْسٌ “heads” is written روس , for rȗ.suṋ. In all of these examples the long vowel V₁V₁ < (V₁)V₁.ʔV₁ or V₁ʔ.V₁(V₁) must have been at first pronounced with a two-peak stress (which would have made it feel like two identical vowels in a hiatus).
§η. A hiatus between a high vowel (i or u) and a low vowel (a) is filled by a secondary glide V₁.ʔV₂ > V₁.V₂ > V₁.wV₂ or V₁.yV₂. For example, the word ya.ʔi.sa يَئِسَ “(he) despaired” is written ييس , for ya.i.sa > ya.yi.sa; yaw.ma-ʔi.ðin يَوْمَئِذٍ “that day” is written يوميذ , for yaw.ma-i.ðin > yaw.ma-yi.ðin; la-tu.nab.ba.ʔun.na لَتُنَبَّؤُنَّ “you(masc. plur.) will be told” is written لتنبون , for la-tu.nab.ba.un.na > la-tu.nab.ba.wun.na; ba.ʔȋ.suṋ بَئِيْسٌ “miserable” is written بيس , for ba.ȋ.suṋ > ba.yȋ.suṋ; ra.ʔȗ.fuṋ رَؤُوْفٌ “compassionate” is written روف , for ra.ȗ.fuṋ > ra.wȗ.fuṋ; sȃ.ʔi.luṋ سَاْئِلٌ “asking (person)” is written سايل , for sȃ.i.luṋ > sȃ.yi.luṋ; ni.sȃ.ʔu.kum نِسَاْؤُكُمْ “your(masc. plur.) women” is written نساوكم , for ni.sȃ.u.kum > ni.sȃ.wu.kum; fi.ʔa.·tuṋ فِئَةٌ “faction” is written فية , for fi.a.·tuṋ > fi.ya.·tuṋ; ri.ʔȃ.ʔa رِئَاْءَ “hypocrisy(construct acc.)” is written ريا , for ri.ȃ > ri.yȃ; yu.ʔax.xi.ru يُؤَخِّرُ “(he) delays/will delay” is written يوخر , for yu.ax.xi.ru > yu.wax.xi.ru; and su.ʔȃ.luṋ سُؤَاْلٌ “asking” is written سوال , for su.ȃ.luṋ > su.wȃ.luṋ.
§θ. A hiatus between two different high vowels (i.u or u.i) is filled by the secondary glide y, and next it is treated in accordance with §II.21.A.h.ν.. For example, ʔa–ʔu.nab.bi.ʔu.kum أَأُنَبِّئُكُمْ “should I inform you(masc. plur.)?” is written أونبيكم , for ʔa-u.nab.bi.u.kum > ʔa-wu.nab.bi.yu.kum > ʔa-wu.nab.bȋ.kum; nab.bi.ʔȗ-nȋ نَبِّئُوْنِيْ “inform [you](masc. plur.) me” is written نبوني , for nab.bi.ȗ-nȋ > nab.bi.yȗ-nȋ > nab.bȗ-nȋ; mus.tah.zi.ʔȗ.na مُسْتَهْزِئُوْنَ “scoffers(masc. plur.)” is written مستهزون , for mus.tah.zi.ȗ.na > mus.tah.zi.yȗ.na > mus.tah.zȗ.na.
§ι. According to Sibawayh (chapter 497), “some Arabs” said at pause ʔ·al-waθw اَلْوَثْوْ , ʔ·al-waθy اَلْوَثْيْ , and ʔ·al-waθȃ اَلْوَثَاْ for standard ʔ·al-waθʔu اَلْوَثْءُ “contusion(nom.),” ʔ·al-waθʔi اَلْوَثْءِ “contusion(gen.),” and ʔ·al-waθʔa اَلْوَثْءَ “contusion(acc.).” Those forms probably belonged to westerm dialects that elided the hamza·t but preserved the terminal case-vowels as pause. The development was ʔ·al-waθʔū > ʔ·al-waθū > ʔ·al-waθu, ʔ·al-waθʔī > ʔ·al-waθī > ʔ·al-waθi, and ʔ·al-waθʔā > ʔ·al-waθȃ. According to اَلْأَسْتَرآبَاْدِيُّ (Chaim Rabin, p. 134, §r), the word ʔ·ar-ridʔu اَلْرِّدْءُ “assistance(nom.)” was pronounced “in Ħiǵȃz” ʔ·ar-ridu at junction and ʔ·ar-rid at pause. This pronunciation probably existed in the Meccan dialect where the terminal nominative and genitnive vowels were lost at pause. In the Qurʔȃn the word difʔuṋ دِفْءٌ “warmth(nom.)” is written دف , perhaps for the pronunciation difuṋ. The word ǵuzʔuṋ جُزْءٌ “warmth(nom.)” is written جز , perhaps for the pronunciation ǵuzuṋ. The word ridʔaṋ رِدْءًا “assistance(acc.)” is written ردا , perhaps for the pronunciation ridaṋ (that is, ridan at junction and ridȃ at pause). The word ʔ·al-marʔu اَلْمَرْءُ “the man” is written المر , perhaps for the pronunciation ʔ·al-maru. Some Qurʔȃn readers read ǵuzzun and ʔ·al-marru at junction. Rabin (p. 134, §t) thinks that these readings were attempts to adapt the biradical words *ǵuzun and *ʔ·al-maru to the triradical form C₁VC₂C₂–.
§κ. According to Sibawayh (chapter 497), “those who pronounced a hamza·t” (اَلَّذِيْنَ يُحَقِّقُوْنَ اَلْهَمْزَةَ) said at pause ʔ·l-kalaw اَلْكَلَوْ , ʔ·l-kalay اَلْكَلَيْ , and ʔ·l-kalȃ اَلْكَلَاْ for standard ʔ·l-kalaʔu اَلْكَلَأُ “herbage(nom.),” ʔ·l-kalaʔi اَلْكَلَأِ “herbage(gen.),” and ʔ·l-kalaʔa اَلْكَلَأَ “herbage(acc.).” “Those who did not pronounce a hamza·t (among) the people of Ħiǵȃz” (اَلَّذِيْنَ لا يُحَقِّقُوْنَ اَلْهَمْزَةَ مِنْ أَهْلِ اَلْحِجَاْزِ) said at pause ʔ·l-xabȃ اَلْخَبَاْ for all the three cases of ʔ·l-xabaʔu اَلْخَبَأُ “hidden thing?” (I am not sure how correct the word ʔ·l-xabaʔu اَلْخَبَأُ is, but الزمخشري mentioned the same information with the example ʔ·l-kalaʔu اَلْكَلَأُ “herbage;” see المفصل p. 161 and Rabin p. 141).
So according to Sibawayh and الزمخشري , some people said at pause ʔ·l-kalaw(nom.), ʔ·l-kalay(gen.), and ʔ·l-kalȃ(acc.), while others said always ʔ·l-kalȃ(nom./gen./acc.). Sibawayh connected this variation to the elision of the hamza·t, but to me it seems that the hamza·t is elided in all of these examples. What he meant seems to be that some speakers replaced the hamza·t with a glide, and this glide was not elided at pause, and so they “pronounced the hamza·t” at pause. In fact, this is a very important note from Sibawayh, because it helps us understand the variation in Qurʔȃnic spelling of words that end with –aʔV. For example, the word ʔ·l-malaʔu اَلْمَلَأُ “the nobels(nom.)” is written in the Qurʔȃn in two ways الملوا and الملا , the word nabaʔuṋ نَبَأٌ “news(nom.)” is written نبوا and نبا , the word nabaʔiṋ نَبَأٍ “news(gen.)” is written نباي (in sȗra·t VI:34) and نبا , and the word ʔatawakkaʔu أَتَوَكَّأُ “(I) lean(indicative)” is written اتوكوا . A similar variation is seen in the spelling of words that end with *–āwV and *–āyV. For example, the word *ǵazāyuṋ > ǵazȃʔuṋ جَزَاْءٌ “retribution(nom.)” is written جزوا and جزا , the word ʔanbȃʔu أَنْبَاْءُ “news(nom. plur.)” is written انبوا and انبا , the word šurakȃʔu شُرَكَاْءُ “partners(nom. plur.)” is written شركوا and شركا , the word ḍuʕafȃʔuṋ ضُعَفَاْءُ “weak(nom. plur.)” is written ضعفوا and ضعفا , and the word ʔȃnȃʔi آنَاْءِ “periods(construct gen. plur.)” is written اناي .
The spelling نباي for nabaʔiṋ نَبَأٍ “news(gen.)” should be compared to افاين for *ʔa-fa-yin < ʔa-fa-ʔin أَفَإِنْ “if then,” and to شاي for *šay(y)uṋ < šayʔuṋ شَيْءٌ “thing.” It seems that ــاي can mean ay, so نباي probably means *nabay. This is identical to the pausal form ʔ·l-kalay اَلْكَلَيْ “herbage(gen.)” which Sibawayh attributed to “those who pronounce the hamza·t.” The variant spelling نبا represents *nabȃ, which is according to Sibawayh the pausal form of “those who do not pronounce the hamza·t.”
Following is an interesting story cited in the Lisȃn:
وَفِي حَدِيثِ اِبْنِ عَبَّاسٍ رَضِيَ اللهُ عَنْهُمَا أَنَّهُ سُئِلَ عَنْ قَتْلِ الْمُحْرِمِ الْحَيَّاتِ، فَقَالَ “لَا بَأْسَ بِقَتْلِه الْأَفْعَوْ وَلَا بَأْسَ بِقَتْلِ الْحِدَوْ”، فَقَلَبَ الْأَلِفَ فِيهِمَا وَاوًا فِي لُغَتِهِ، أَرَادَ الْأَفعَى، وَهِيَ لُغَةُ أَهْلِ الْحِجَازِ. قَالَ اِبْنُ الْأَثِيرِ “وَمِنْهُمْ مَنْ يَقْلِبُ الْأَلِفَ يَاءً فِي الْوَقْفِ، وَبَعْضُهُمْ يُشّدِّدُ الْوَاوَ وَالْيَاءَ، وَهَمْزَتُهَا زَائِدَةٌ” [لسان العرب “فعا”]
وَرُوِيَ عَنْ اِبْنِ عَبَّاسٍ أَنَّهُ قَالَ “لَا بَأْسَ بِقَتْلِ الْحِدَوْ وَالْإِفْعَوْ لِلْمُحرِمِ”، وَكَأَنَّهَا لُغَةٌ فِي الْحِدَأِ [لسان العرب “حدأ”]
It is told that ʔIbn ʕAbbȃs (Prophet Muhammad’s cousin who died c. 687 AD) said ʔafʕaw أَفْعَوْ or ʔifʕaw إِفْعَوْ for ʔafʕa·y أَفْعَىْ “snake,” and ħidaw حِدَوْ for ħidaʔuṋ حِدَأٌ “kite, accipitrid.” This story does not seem very accurate, because the form ʔafʕaw أَفْعَوْ was said by Sibawayh to exist among some of the Ṭayyiʔ طَيِّئٌ of northwestern Naǵd (§II.21.A.h.η.VII.). As far as I know, word stems with more than 3 letters never end with ـــا or ـــوا in the Qurʔȃn (except when there is a yȃʔ before the last letter, see §II.21.A.h.κ.). The word ʔafʕ·y أَفْعَىْ in Qurʔȃnic orthography would have been افعى*. Since words with terminal *–aw are almost always written in the Qurʔȃn with terminal ـــا , it seems unlikely that terminal ــى in افعى* would have been pronounced **–ay in Mecca. More likely, it would have been pronounced *–e (§II.21.A.h.η.). Another possible inaccuracy in the saying attributed to ʔIbn ʕAbbȃs is the word بِقَتْلِه “in (his) killing,” which could stand for *bi-qatlih < *bi-qatlī, that is, the terminal –h could be “the hȃʔ of silence” هَاْءُ الْسَّكْتِ (§II.19.C.a.). If so, this is inaccurate, because the vowel of the genitive case was elided at pause in Mecca, as can be deduced from Qurʔȃnic orthography (§II.19.B.).
It is possible that the statement attributed to ʔIbn ʕAbbȃs is a parody on Ħiǵȃzi speech in general (there are several animals which a pilgrim is allowed to kill in Mecca, why is it that only two rhyming animal names are mentioned in this statement ?) However, the word ħidaw حِدَوْ for ħidaʔuṋ حِدَأٌ might indeed be a Meccan word, since the Qurʔȃn has many similar forms.
It seems that in Mecca the ending *–a.V < *–a.ʔV was pronounced in two alternative ways. The first way was to add a secondary glide when the terminal vowel was u or i, so that the ending became *–a.wV/*–a.yV. At pause this was reduced to *–aw/*–ay, hence the Qurʔȃnic spelling نبوا for *nabaw < *nabawun < *nabaun < *nabaʔun نَبَأٌ “news(nom.),” and نباي for *nabay < *nabayin < *nabain < *nabaʔin نَبَأٍ “news(gen.).” The other way of pronouncing the ending *–a.V < *–a.ʔV was to keep the hiatus (the “intermediate hamza·t” اَلْهَمْزَةُ اَلَّتِيْ بَيْنَ بَيْنَ ). At pause this was reduced to *–a, hence the Qurʔȃnic spelling نبا for *naba < *nabaun < *nabaʔun نَبَأٌ “news(nom.),” and نبا for *naba < *nabain < *nabaʔin نَبَأٍ “news(gen.).”
The ending *–ȃ.V < *–ȃ.wV/*–ȃ.yV was treated similarly. This ending had two pausal variants when the terminal vowel was u or i, the first was *–ȃw/–ȃy and the second was *–ȃ.
§λ. Words with terminal *–ȃ.VV < *–ȃ.ʔVV/*–ȃ.wVV/*–ȃ.yVV were reduced at pause to *–w(V)/*–y(V) when the terminal long vowel was high, e.g. the word ǵȃ.ʔȗ جَاْؤُوْاْ “(they)(masc. plur.) came” is written in the Qurʔȃn جاو , perhaps for *ǵȃwu < *ǵȃ.wȗ < *ǵȃ.ȗ < *ǵȃ.ʔȗ, the word wa.rȃ.ʔȋ وَرَاْئِيْ “behind me” is written in the Qurʔȃn وراي , perhaps for *wa.rȃ.yi < *wa.rȃ.yȋ < *wa.rȃ.ȋ < *wa.rȃ.ʔȋ (see Chaim Rabin, p. 139, §dd).
§μ. In the modern spoken dialects of Arabic the hamza·t sound is often elided, although not as regularly as in Classical west Arabian. A syllable-terminal hamza·t is usually dropped with compensatory lengthening of the preceding vowel (unless it is a terminal vowel in which case it is pronounced short). Thus, in the dialect of Aleppo rās for Classical raʔsuṋ رَأْسٌ “head,” yāxod and yākol for yaʔxuðu يَأْخُذُ “(he) takes/will take” and yaʔkulu يَأْكُلُ “(he) eats/will eat” (but maʔxūd for maʔxȗðuṋ مَأْخُوْذٌ “taken,” maʔkūl for maʔkȗluṋ مَأْكُوْلٌ “eaten,” and taʔxīr for taʔxȋruṋ تَأْخِيْرٌ “making late”), bīr for biʔruṋ بِئْرٌ “well,” lūlu for luʔluʔuṋ لُؤْلُؤٌ “pearls,” and sama for samȃʔuṋ سَمَاْءٌ “heaven.” A syllable-initial hamza·t is dropped in mara·t for marʔa·tuṋ مَرْأَةٌ “woman” and rūs for ruʔȗsuṋ رُؤُوْسٌ “heads” (but saʔal for saʔala سَأَلَ “(he) asked,” sēʔel for sȃʔiluṋ سَاْئِلٌ “asking,” tʔaxxar for taʔaxxara تَأَخَّرَ “(he) was/became late,” and mʔaxxer for muʔaxxiruṋ مُؤَخِّرٌ “making late(active participle)“). Verbs III=ʔ are merged with verbs III=w/y (this merger probably existed in Classical west Arabian. I will talk about it in detail when I talk about verbs). Thus, Classical qaraʔa قَرَأَ “(he) read” inflicts in Aleppine Arabic just like Classical saʕa·y سَعَىْ “[he] sought, moved,” and Classical badaʔa بَدَأَ “[he] started” inflicts in Aleppine Arabic just like Classical baqiya بَقِيَ “[he] remained.”
۩ DROPPED HAMZA·T IN THE STANDARD LANGUAGE
§ν. In perfect Classical Arabic (the ‘Clear Language’ اَلْلُّغَةُ اَلْفَصِيْحَةُ ) the hamza·t should not be dropped, except in syllables of the form ʔVʔ which regularly become ʔVV.
|Word with dropped hamza·t||Original Form (Not Used)||Meaning|
|أُوْمِنُ||أُؤْمِنُ||“(I) believe in (intr.)”|
However, in the actual lexicon of perfect Classical Arabic some words with dropped hamza·t’s were used regularly. For example, the word Šȃmuṋ شَاْمٌ , which is the Classical Arabic equivalent of the word “Syria,” is originally Šaʔmuṋ شَأْمٌ . The form with the hamza·t was used in Classical Arabic, but not as commonly as the form with dropped hamza·t. (The word Šaʔmuṋ شَأْمٌ means “left, north,” thus it is an antonym of Yamanu يَمَنُ “Yemen” which means “right, south.”) Other examples are tȃrȋxuṋ تَاْرِيْخٌ “history” which is altered from taʔrȋxuṋ تَأْرِيْخٌ “dating” (the root is ʔrx), nabȋyuṋ = nabiyyuṋ نَبِيٌّ “prophet” which is altered from nabȋʔuṋ نَبِيْءٌ (the root is nbʔ), rayyisuṋ رَيِّسٌ “chief” which is altered (per §II.21.A.h.ζ.II.) from *rayȋsuṋ < *ra.ȋsuṋ < ra.ʔȋsuṋ رَئِيْسٌ (the root is rʔs), and bidȃya·tuṋ بِدَاْيَةٌ “beginning” which is altered from bidȃʔa·tuṋ بِدَاْءَةٌ (the root is bdʔ) by the change of the root from III= ʔ to III= y.
The perfect Classical word yarȃ يَرَىْ “(he) sees/will see” is altered from yarʔȃ يَرْأَىْ < *yarʔay يَرْأَيْ* (the root is rʔy, c.f. Hebrew yirʔɛh יִרְאֶה < *yirʔay יִרְאַי*). The word malȃkuṋ/malakuṋ مَلَاْكٌ/مَلَكٌ “messenger, angel” seems to be cognate with milʔakuṋ مِلْأَكٌ with the same meaning (c.f. Hebrew malʔāx מַלְאָךְ < *malʔak– מַלְאַךּ* “messenger, angel”). Apparently the original root was ʔlk “convey” (c.f. hlk) but was metathesized to lʔk. The Classical philologists were aware of this etymology:
قَالَ الْلَّيْثُ [بِنُ الْمُظَفَّرِ]: الْمَلَكُ وَاحِدُ الْمَلَائِكَةِ إِنَّمَا هُوَ تَخْفِيفُ الْمَلْأَكِ، وَاِجْتَمَعُوا عَلَى حَذْفِ هَمْزِهِ، وَهُوَ مَفْعَلٌ مِنْ الْأَلُوكِ، وَقَدْ ذَكَرْنَاهُ فِي المُعْتَلِّ […] قَالَ الْكِسَائِيُّ: أَصْلُهُ “مَأْلَكٌ” بِتَقْدِيمِ الهَمْزَةِ، مِنْ “الْأَلُوكِ” وَهِيَ الرِّسَالَةُ، ثُمَّ قُلِبَتْ وَقُدِّمَتْ الْلَّامُ فَقِيلَ “مَلْأَكٌ” [لسان العرب “ملك”].
A similar development may be behind the word šamȃluṋ/šimȃluṋ شمَاْلٌ “left, north” which seems to be cognate with šamʔaluṋ شَمْأَلٌ , šamʔalluṋ شَمْأَلٌّ , and šaʔmaluṋ شَأْمَلٌ “northern wind” (c.f. Hebrew s₂əmȏl שְׂמאֹל < *s₂imʔal– שִׂמְאַל* “left,” Ugaritic s₁mʔl 𐎌𐎎𐎀𐎍“left,” and Ancient South Arabian s₂ʔml “left, north”). The original root was perhaps šʔml < šʔm and this was metathesized to šmʔl. The liquid l may be explained if we assume a dissimilation *s₂aʔmam– > *s₂aʔmal–.
وَالْشَّمَالُ: الْرِّيحُ الْتِّي تَهُبُّ مِنْ نَاحِيَةِ الْقُطْبِ، وَفِيهَا خَمْسُ لُغَاتٍ: شَمْلٌ بالتسكين، وشَمَلٌ بالتحريك، وشَمَالٌ، وشَمْأَلٌ مهموز، وشَأْمَلٌ مقلوب، قال: وربما جاء بتشديد اللام؛ قال الزَّفَيانُ: تَلُفُّه نَكْباءُ أَو شَمْأَلُّ […] وقال ابن الأَعرابي: […] وهي الشَّمُولُ، والشَّيْمَلُ، والشَّمْأَلُ، والشَّوْمَلُ، والشَّمْلُ، والشَّمَلُ [لسان العرب “شمل”].
۩ #ʔ– > #w–/#y–
§ξ. A word-initial hamza·t #ʔ– is sometimes changed to #w–. For example, the root ʔħd “one” has a variant wħd which appears in words such as waħaduṋ وَحَدٌ “one” and wȃħiduṋ وَاْحِدٌ “one.” In modern spoken dialects of Egypt we find widn “ear” for Classical ʔuðunuṋ/ʔuðnuṋ أُذنٌ , wȃxid “taking” for Classical ʔȃxiðuṋ آخِذٌ , and wȃkil “eating” for Classical ʔȃkiluṋ آكِلٌ .
In some cases the change #ʔ– > #w– could have started from imperfective Form IV verbs with elided ʔ. For example, the root wkd, which is a variant of ʔkd “confirm; emphasize,” could have developed from yuʔakkidu يُؤَكِّدُ > yuakkidu > yuwakkidu يُوَكِّدُ “(he) emphasizes/will emphasize.” The same development could be behind ʔṣd/wṣd “close” and ʔkf/wkf “donkey saddle” (see Chaim Rabin, p. 141, §ff.). However, it is hard to imagine that all of the changes #ʔ– > #w– developed in this manner. It is possible that #ʔ– was sometimes dropped and then an on-glide developed in its place, e.g. the word ʔaħaduṋ أَحَدٌ “one” could have developed as follows ʔaħaduṋ > **aħaduṋ > *ʷaħaduṋ > waħaduṋ وَحَدٌ (this development perhaps presupposes that words of the form CV.CV.CV were not accented on the first syllable).
Modern examples of the change #ʔ– > #y– include the Bedouinic yisȋr for Classical ʔasȋruṋ أَسِيْرٌ “captive.” In some Syrian dialects one can hear yalli as a variant of ʔalli/ʔilli for the relative pronoun. In Lebanon the demonstratives ʔaħħuwwe and ʔaħħiyye have variants yaħħuwwe and yaħħiyye. In Classical Arabic the root ʔmm “head to, direct to” has a variant ymm.
۞ SIDE NOTE
In my opinion, the word ʔ·Al-Yamȃma·tu الْيَمَاْمَةُ , which is the name of a Classical eastern Arabian region, is originally synonymous with ʔamȃmuṋ/yamȃmuṋ أَمَاْمٌ/يَمَاْمٌ “front, east;” c.f. the name of Greater Syria ʔ·Aš-Šaʔmu الشَّأْمُ “left, north” and the name of Yemen ʔ·Al-Yamanu الَيَمَنُ “right, south.”
The root ʔmn “safety” is perhaps etymologically related to ymn “well-being.” The word yaduṋ يَدٌ “hand” is probably cognate with the root ʔdy “hand, deliver;” this root is found, for example, in the word *ʔadayatun > ʔadȃ·tuṋ أَدَاْةٌ “tool.”
The preformative element ʔa– is sometimes changed to ya–, as can be seen in the following pairs:
|أَحْمَدُ||يَحْمَدُ||male personal name|
|أَزِيْدُ||يَزِيْدُ||male personal name|
|أَسْلَمُ||يَسْلَمُ||male personal name|
|أَعْلَىْ||يَعْلَىْ||male personal name|
|أَثْرِبُ||يَثْرِبُ||Former name of Medina|
|أَلَمْلَمُ||يَلَمْلَمُ||Mountain in western Arabia|
The proper nouns with initial ya– were taken by old linguists (and also by modern ones) to be verbal in origin, because they look similar to verbal stems with initial preformative ya–, but it is likely that ya– was originally a mere phonetic variant of ʔa–, and the proper nouns with initial ya– were misconstrued as verbal forms because of the chance resemblance to verbal stems with initial ya–.
۞ SIDE NOTE
The preformative element yV– which appears in 3rd person verbal stems has often been considered pronominal in origin, but it is likely that this is just a variation of ʔV– which appears in the 1st person singular stem. Originally these prefixes were not pronouns. They were reanalyzed as pronouns by pre-Proto-Semitic speakers and they were modified as such.
The Ancient Egyptian pronouns ı͗nk “I” and ı͗nn “we” correspond to Classical Arabic ʔanȃ أَنَاْ “I” and naħnu نَحْنُ “we” and to Akkadian anāku “I” and nīnu < *niħnu “we.” Antonio Loprieno (in Ancient Egyptian: A linguistic introduction (1995)) reconstructs the Ancient Egyptian pronouns as *yanak (c.f. Coptic anok ⲁⲛⲟⲕ) and *yanan (c.f. Coptic anon ⲁⲛⲟⲛ). Loprieno believes that ı͗ was pronounced y in the Egyptian of the Old Kingdom, which sounds convincing given that the Egyptian of the Old Kingdom lacked another character for y (the character ı͗ı͗ = y was used in the Middle Kingdom, after the sound of ı͗ changed to ʔ). So it appears that the element *ʔan– of the personal pronouns was changed in Ancient Egyptian to *yan–.
Perhaps the change #ʔ– > #y– happened by dropping the glottal stop and replacing it by an on-glide, i.e. #ʔ– > #∅– > #ʸ– > #y–.