Compendious Syriac Grammar/Part 1/Chapter 2

2804460Compendious Syriac Grammar — I. 2. PhonologyJames Alexander CrichtonTheodor Nöldeke

II. PHONOLOGY

1. CONSONANTS.

GENERAL STATEMENT.

Beginning of the syllable. § 20. Every word and every syllable commences with a consonant. That no word can begin with a vowel sound is expressed clearly in Semitic writing by ܐ [preceding such sound], e. g. ܐܴܬܷܐ āthē, or rather ʾāthē "comes"; ܐܾܘܪܚܳܐ ʾurḥā "a way"; ܐܻܝܕܳܐ ʾīδā "hand," &c. In cases like ܝܺܕܰܥ "knew", the word is spoken as if it stood ܐܻܝܕܰܥ ʾīδaʿ, and so it is even written at times (§ 40 C).

No Syriac word begins originally with a double consonant. Yet such a consonant seems to have been produced by the falling away of a very short vowel in ܫܬܴ݁ܐ, ܫܬܻ݁ܝܢ štā, štīn (as well as ܐܷܫܬܴ݁ܐ, ܐܷܫܬ݁ܝܢ) "six", "sixty" (in East-Syriac also, ܫܬܻ݁ܬܴܝܳܐ "the sixth"; cf. the forms for sixteen § 148 B); in the later pronunciation still oftener, and even in other cases, as perhaps in ܟܣܷܐ ksē from ke̊sē "covered".

Doubling. § 21. The West-Syrians appear to have lost long ago the original doubling of a consonant; the East-Syrians seem generally to have retained it: the former, for example, pronounce עַמָּא "people", ܥܱܡܳܐ ʿamō, the latter ܥܲܡܵܐ ʿammā. Nearly every consonant then is to be held as doubled, which is preceded by a short vowel and followed by any vowel, thus ܩܱܛܶܠ "murdered", ܢܷܣܱܒ "takes" are pronounced qaṭṭel, nessav.

The absence of doubling may be relied on only when a softened consonant continues soft, e. g. ܐܷܬ݂ܳܐ ʾethā "came", not ʾeththā, for this softening, or assibilation, is inadmissible in a doubled letter; while on the contrary the hard sound in such a consonant after a vowel is a sure token of doubling, e. g. ܢܱܦܝܩ nappīq "gone forth". How far the gutturals ܥ‍ and ܗ underwent a real doubling is a matter of question; but the treatment of the vocalisation for the most part is the same as if such doubling had occurred (cf. Hebr. בִעַר,‎ מַהֵר). The case is similar with ܪ, which also the East-Syrians at a pretty early date had already ceased to double, but for which they occasionally at least turned a foregoing a into ā.

In many cases the doubling has entered in a secondary way, as in ܐܠܗܐ allāhā "God", ܐܕ݁ܒܚ eddabbaḥ "I sacrifice".

B. The doubling at all events very early fell away, when merely a sheva followed the doubled consonant, e. g. in ܪܷܓ݁ܬ݂ܳܐ "desire", properly regge̊thā, then regthā, and even very early through assimilation (§ 22) rekthā so ܒܷܙܬ݂ܳܐ bezze̊thā "booty", bezthā, besthā. Thus ܡܼܬܓ݁ܫ̣ܫܵܐ "it is touched", properly methgašše̊šā, was early pronounced like methgaššā or even methgašā.

C. A very ancient dissolving of the doubling in the case of r, with compensation in lengthening the vowel, appears to occur in ܓܶܐܪܴܐ gērā "arrow" from garrā; ܚܹܐܪ̈ܶܐ ḥērē (ḥērīn &c.) "free", from ḥarrē; ܒܹܪ̈ܝܴܬܴܐ bēryāthā "streets" from barryāthā. Thus perhaps also ܨܶܝܕ݂ (ܨܶܝ̈ܕ݂ܰܗ̄ܝ &c.) "with" from ṣadd.

D. Consonants written double were originally separated by a vowel, though very short, e. g. ܣܡܡ̈ܐ φάρμακα samåmē, later sammē, ܓܠ̈ܠܐ "waves" galålē, later gallē ܓܕ̈ܕܐ "wormwood" gedådē, later geddē. By a false analogy even ܣܡ̈ܡܢܐ φάρμακα sammānē is accordingly often written instead of ܣܡ̈ܢܐ, and in fact ܣܡܡܐ for the singular instead of ܣܡܐ sammā and similarly in like cases. An actual exception to that rule is furnished only by cases like ܐܬܬܣܝܡ or ܐܬܣܝܡ ette̊sīm "was set"; ܐܬܬܥܝܪ or ܐܬܥܝܪ ette̊ʿīr "was awakened" &c. (§§ 36. 177 B).

In Greek words letters are sometimes written double, even when such doubling does not occur in the original, e. g. ܦܝܠܠܝܦܘܣ Φίλιππος often instead of ܦܝܠܝܦܘܣ or ܦܝܠܝܦܦܘܣ.

Assimilation. § 22. When two consonants came together in the living speech, and still more in the somewhat artificial recitation of the Bible in religious service, the first consonant was frequently modified by the second, so that a media before a tenuis was turned into a tenuis, a tenuis before a media into a media, and so forth. ܙܬ was pronounced like ܣܬ (e. g. ݂ܪܱܓܽܘܙܬ݂ܳܢܳܐ "vehemently angry" like ܪܱܓܽܘܣܬ݂ܳܣܳܐ), for ܙ is a media and ܣ a tenuis like ܬ݂ (in spite of the assibilation); ܙܟ‍ like ܣܟ‍ (e. g. ܢܷܙܟܷ݁ܐ "conquers" like ܢܷܣܟܷ݁ܐ; ܕܰܙܟ݂ܰܪܝܳܐ "of Zacharias" like ܕܰܣܟ݂ܰܪܝܳܐ); vice versa ܣܕ like ܙܕ (e. g. ܚܶܣܕܳ݁ܐ "disgrace" like ܚܶܙܕܳ݁ܐ). Farther ܓܬ݂ was given like ܟܬ݂ (e. g. ܪܱܝܽܘܓ݂ܬ݂ܳܢܳܐ "greedy" like ܪܱܝܽܘܟ݂ܬ݂ܳܢܳܐ), and even ܩܬ݂, with suppression of the emphasis before the last unemphatic ܬ, like ܟ݁ܬ݂ (e. g. ܥܱܝܽܘܩܬ݂ܳܢܳܐ "sorrowful" like ܥܱܝܽܘܟ݁ܬܳܢܳܐ. The East-Syrians went much farther in this process, for they prescribed e. g. ܠܡܷܕ݂ܒܱ݁ܪ even for ܠܡܷܬ݂ܒܱ݁ܪ "to break"; ܢܹܐܓ݁ܕ݁ܘܰܢ for ܢܹܐܩܕ݁ܘܰܢ; and they gave to ܫ‍ immediately before ܒ‍, ܓ, ܕ, the sound of the French j, ge (Pers. ژ‎), e. g. in ܚܰܘܫܒܳܢܳܐ "an account". This subject might be treated at great length. Notice that such assimilations take place even when the consonants affected were originally separated by a sheva ().—The written language exhibits only a few traces of these changes.[1]

Rem. A very ancient reversed assimilation consists in ܩܬ always becoming ܩܛ‍ in Aramaic roots[2] at the beginning of the word, as the emphatic ܛ‍ corresponds more accurately to ܩ‍ than does ܬ. Similar equalisations in all roots might farther be pointed out.

Rukkākhā and Quššāyā. R. and Q. in individual words.

RUKKĀKHĀ AND QUŠŠĀYĀ.

§ 23. A. The rules for Rukkākhā, i.e. the soft (assibilated, hissing, or aspirated) pronunciation and for Quššāyā, i.e. the hard (or unaspirated) pronunciation, originally affect all the letters ܬ ܦ‍ ܟ‍ ܕ ܓ ܒ‍ [Beghadhkephath] in equal measure. But the East-Syrians for a very long time have nearly always given ܦ‍ a hard sound; only in the end of a syllable have they sometimes given it a soft pronunciation.[3] The following rules accordingly are not applicable to the East-Syrian pronunciation of ܦ‍.

B. These letters are hard in the beginning of words, e. g. ܒܱ݁ܝܬܴ݁ܠ "house", ܓܱ݁ܡܠܴܐ "camel," &c. (but notice § 24 and § 25).

C. ܬ ܦ‍ ܟ‍ ܕ ܓ ܒ‍ experience R.—(i. e. take the soft pronunciation) after any vowel, however short, when they do not happen to be doubled. Thus after a full vowel ܩܳܒ݂ܶܠ, ܣܺܝܓ݂, ܫܱܦ݂ܠܱܐ, ܩܽܘܕ݂ܫܳܐ, ܣܷܬ݂ܪܴܐ, ܣܱܟ݂ܠܴܐ &c.

On the other hand these letters undergo Q. (i. e. take the hard form) when they are doubled: ܩܱܒܷ݁ܠ (קַבֵּל), ܣܱܓܻ݁ܝ (סַגִּי), ܡܩܱܕܰ݁ܫ, ܣܽܘܟܴ݁ܠܴܐ, ܫܱܦ݁ܠܷܬ, ܣܷܬܴ݁ܪܴܐ, &c., and even after long vowels ܪܴܓܻ݁ܝܢ (rāggīn "they desire") ܒ݁ܬܷ݁ܐ (bāttē "houses"), &c.

Exception[errata 1]: ܐܱܝܟ݂ "as", which is pronounced akh.

ܶEven the mere sheva mobile effects R. just as a vowel would: ܩܒ݂ܘܿܠ (qe̊vol), ܣܓ݂ܳܐ, ܡܱܡܠܟ݂ܺܝܢ, &c. Thus is it also when one of the particles ܘ ܕ ܠ‍ ܒ‍ is prefixed: ܒܴ݁ܢܶܐ; but ܠܒ݂ܳܢܶܐ (le̊vānē); ܓ݁ܐܷܪܴܐ, but ܒܓ݂ܶܐܪܴܐ, &c. So too is it when several of these words or particles are prefixed, e. g. ܓܱ݁ܪ̈ܡܶܐ: ܒ݁ܓܱܪ̈ܡܶܐ, ܕܱ݁ܒ݂ܓ݂ܰܪ̈ܡܷܐ, ܘܕ݂ܰܓ݂ܰܪ̈ܡܷܐ; ܒܱ݁ܝܬܴ݁ܐ: ܘܱܒ݂ܕ݂ܰܠܒ݂ܰܝܬܴ݁ܐ, &c. Except upon the first consonant, these prefixes however have no effect, thus, ܟ݁ܐ݂ܳܒ݂ܳܐ ke̊thāvā, ܠܟ݂ܬ݂ܳܒ݂ܳܐ lakhthāvā, originally lakhe̊thāvā, not lakhtāvā &c.

Regularly the sheva mobile has a softening effect after a consonant originally doubled, thus ܡܚܰܫܒ݂ܺܝܢ (מְחַשְּׁבִֿין), ܪܷܓ݁ܬ݂ܳܐ (רֶגְּתָֿא), ܡܱܚܬ݂ܳܐ maḥḥe̊thā = manḥe̊thā), &c. So also, of course, when the consonant furnished with sheva mobile is preceded by another which is quite vowelless, as in ܡܱܣܗܕ݂ܺܝܢ, ܕ݂ܶܚܠܬ݂ܳܐ, ܡܱܡܠܟ݂ܳܐ, &c.

D. But many a sheva mobile fell away (sheva mobile transmuted into sheva quiescens) at a time when the influence which it exercised upon the softening process (Rukkākhā) was still a living one, with the result that the influence of the hardening process (Quāyā) in turn appeared. On the other hand such falling away occasionally came about at a time when the influence referred to was no longer in being, so that Rukkākhā remained effective even after the disappearance of sheva mobile. Upon the whole R. has been abandoned more completely in the case of the falling away of an that had originated from i (e), than in that of an from a: compare ܓܰܪܒ݂ܳܐ "scabies" from garăvā, with ܓܰܪܒܴ݁ܐ "scabiosus" from garĭvā. It makes no difference whether the foregoing syllable,—now a closed one (ending in sheva quiescens),—has a long or a short vowel; cf. ܢܳܣܒܻ݁ܝܢ, ܩܳܬ݂ܒܻ݁ܝܢ, ܦ݂ܠܓܾ݁ܘܬ݂ܳܐ, and other derivatives from the act. part. Peal[4]; ܫܩܽܘܠܬܴ݁ܐ, ܐܱܥܸܝܪܬܷ݁ܗ "I awakened him", &c.

In the interior of words R., when it comes after an earlier sheva mobile unpreceded by two consonants without a full vowel or by a double consonant, is now kept up only here and there, and that particularly in the verb: cf. even cases like ܢܹܐ̈ܠܕ݁ܵܢ nēldān (nīldōn) "they bring forth children", from nēli̊δān. For the substantive,—cf. cases like ܡܱ̈ܠܟܱ݁ܝ, contrasted with the Hebr. מַלְכֵֿי from malåkhai (but v. § 93) and ܡܱܠܟܾ݁ܘܬ, contrasted with מַלְכֿוּת.

E. The usage in the case of Fem. ܬܴܐ is specially fluctuating, for the ܬ here is often hard after a consonant, and often on the other hand soft. This ܬ has nearly always Q. [i. e. it is pronounced hard, as if with Dag. lene] after syllables which have a long vowel, particularly ī or ū, e. g. ܫܱܪܻܝܪܬܴ݁ܐ, ܩܱܕܺܝܫܬܴ݁ܐ, ܒܺܝܫܬܴ݁ܐ, ܒܺܝܕܬܴ݁ܐ, ܦܪܻܝܩܬܴ݁ܐ, ܕܹܐܪܬܴ݁ܐ, ܚܹܐܪܬܴ݁ܐ, ܢܦܹܐܣܬܴ݁ܐ; ܩܒܽܘܪܬܴ݁ܐ, ܕܪܶܘܟ݂ܬܴ݁ܐ, ܨܽܘܪܬܴ݁ܐ, ܒܬܽܘܠܬܴ݁ܐ, &c.; Exceptions:—ܢܝܳܚܬܴ݁ܐ, ܣܝܳܡܬܴ݁ܐ, ܒܹܥܬ݂ܳܐ[5]; ܪܚܽܘܡܬ݂ܳܐ, and some others. With ā: ܢܝܳܚܬܴ݁ܐ, ܣܝܳܡܬܴ݁ܐ, ܣܱܝܒܳܪܬܴ݁ܐ, ܚܳܠܬܴ݁ܐ, &c.; but ܪܴܡܬ݂ܳܐ, ܡܪܴܡܬ݂ܳܐ, ܡܙܳܥܬ݂ܳܐ, ܜܳܒܬ݂ܳܐ, ܣܴܒܬ݂ܳܐ, ܫܳܥܬ݂ܳܐ, ܥܴܩܬ݂ܳܐ, and a few others. Always Q. (i. e. Quššāyā, or Dag. lene) after ◌ܳ‎‍ܝ, e. g. ܫܳܡܪܴܝܬܴ݁ܐ, ܙܰܢܳܝܬܴ݁ܐ. After syllables with ă, perhaps R. of ܬ somewhat preponderates: ܚܒܱܪܬ݂ܳܐ, ܦܩܱܥܬ݂ܳܐ, ܢܫܱܡܬ݂ܳܐ, ܡܱܘܕܰܥܬ݂ܳܐ, ܡܱܘܗܰܒ݂ܬ݂ܳܐ, ܩܱܪܩܱܦ݂ܬ݂ܳܐ, ܐܷܓܱ݁ܪܬ݂ܳܐ, &c.; yet ܡܱܥܱܠܬܴ݁ܐ, ܫܹܝܫܱܠܬ݂݁ܐ, ܡܫܱܒܱܚܬܴ݁ܐ, ܬܱܚܢܱܢܬܴ݁ܐ, and many others. With ĕ Q. has the preponderance: ܝܺܬ݂ܶܒ݂ܬܴ݁ܐ, ܫܷܒܷ݁ܠܬܴ݁ܐ, ܬܱܠܒܷ݁ܫܬܴ݁ܐ, and many others; yet ܝܺܙܶܦ݂ܬ݂ܳܐ, and so too, forms like ܪܴܚܶܡܬ݂ܳܐ, ܩܳܝܶܡܬ݂ܳܐ (to which ܦ݂ܪܱܚܬ݂ܳܐ, ܥܼܡܱܪܬ݂ܳܐ also belong). So ܚܶܡܬ݂ܳܐ "anger", and the like. With u and o we have ܬܷܫܒܘܿܚܬܴ݁ܐ, ܡܷܐܟܽܘܠܬܴ݁ܐ, ܬܷܐܓ݂ܽܘܪܬܴ݁ܐ, &c., overagainst ܡܫܽܘܚܬ݂ܳܐ, ܐܷܣܟܾ݁ܘܦ݂ܬ݂ܳܐ. Individual peculiarities are very complicated here, and the tradition occasionally varies. On the whole Q. is preferred after r, l, and š, and R. after ʿ, m &c., in the ܬ of the termination ܬܴܐ [i. e. ܬ in that feminine termination, is generally sounded hard after r, l, and š, and R. after ʿ, m &c., in the ܬ of the termination ܬ݂ܐ [i. e. ܬ in that feminine termination, is generally sounded hard after r, l, and š, and soft, or with assibilation, after ʿ and m]. The analogy of words of similar form or meaning has exercised great influence here. Something will be said on this head afterwards in treating of the parts of speech.

F. The quite peculiar Q. of ܫܬܴ݁ܐ, ܫܬܻܳܝܢ (along with ܐܷܫܬܴ݁ܐ, ܐܷܫܬܻ݁ܝܢ) "six", "sixty" points to the loss of a sheva in remote times [v. D].

G. Like ܚܶܡܬ݂ܳܐ "anger" we also have ܚܶܡܬ݂ܝ, ܚܶܡܬ݂ܗܘܿܢ "my, their anger"; here farther, analogy in this way breaks through the old law, that Q. must stand immediately after a consonant [v. C]. Thus ܕܰܗܒ݂ܝ, ܕܰܗܒ݂ܗܘܿܢ "my, their gold", following ܕܰܗܒ݂ܳܐ "gold" (from dahăvā), and many others. Thus the ܬ of the 3. sing. fem. in the Perf. (at least according to the usual pronunciation) remains always soft: ܩܜܰܠܬ݂ܶܗ "she has killed him", ܩܜܰܠܬ݂ܰܢܝ "she has killed me" (as against ܩܜܰܠܬܷ݁ܗ "I have killed him", &c.). On the other hand the ܬ of the 2. pers. in the Perf. is kept hard in all circumstances, thus ܩܜܰܠܬ݁ "thou hast killed" (and ܐܱܢ̄ܬ݁ "thou"), as well as ܓܰܠܻܝܬ݁ "thou hast revealed", ܓܰܠܻܬ݁ܝ "thou (f.) hast revealed"; ܓܰܠܻܝܬ݁ܘܿܢ, ܓܰܠܻܝܬܷ݁ܝܢ "Ye (m. and f.) have revealed" &c.

In other respects too we find remarkable deviations from the fundamental rules, e. g. in ܐܱܪ̈ܒܳܥܼܬ݁ܝܗܶܝܢ (§ 149) "they four (f.)" or "the four of them", where ܬ݂ might have been expected. Although the fundamental rules are still clear, they became practically ineffective even at an early stage; and thus it came about that entirely similar cases often received dissimilar treatment. Besides, fluctuations of all kinds in the dialects and in the school-tradition, manifest themselves in the matter of R. and Q.[6]

H. Original doubling in the termination preserves Q. in ܪܱܒ݁ (like ܪܱܒܴ݁ܐ) "great", ܓܽܘܒ݁ "a pit", ܓܰܒ݁ (ܓܰܒܴ݁ܐ from ܓܰܢܒܴ݁ܐ) "side", ܕܽܘܟ݁ "place"; so too ܐܱܢ̄ܬ݁ at = att from ant "thou"; so also ܠܷܒ݁ܝ leb "my heart" (like ܠܷܒ݁ܐ lebbā), ܓܰܕ݁ܝ gad "my good fortune" (like ܓܰܕ݁ܐ) and the like. On the other hand we have ܫܷܬ݂ "six" (its doubling early disappeared), ܓܶܢ̄ܒ݂ "side" (also ܓܶܢ̄ܒ݂ܝ "my side") and verbal forms like ܫܱܒ݂ "lowered", ܪܱܓ݂ "longed for" (and also in the plural ܪܱܓ݂ܘ &c.).

I. Secondary doubling, which causes Q., we find regularly in the 1st sing. Impf. when the first radical has a vowel, as in ܐܷܕ݁ܽܩܫ "I tread", ܐܷܟܱ݁ܕܶܒ݂ "I tell lies", ܐܷܒܱ݁ܪܟ݂ܳܟ "I bless thee", ܐܷܓܱ݁ܪܷܐ "I hunt," &c. Farther in the Aphel in some verbs middle ܘ: ܐܱܟܻ݁ܝܢ "made ready", ܐܱܟܻ݁ܝܠ "measured", as contrasted with ܐܱܬ݂ܺܝܒ݂ "gave back", &c. (§ 177 D).

J. Words, which are otherwise like-sounding, are often distinguished through R. and Q., as ܓܰܠܻܝܬ݁ "thou hast revealed", and ܓܰܠܻܝܬ݂ "I have revealed"; ܩܷܫܬ݂ܳܐ qešthā from qešše̊thā (f. of Hebr. קַשׁ) "stubble", and ܩܷܫܬܴ݁ܐ (קֶשֶׁת) "a bow", &c.

R. and Q. in closely associated words. § 24. R. appears in the beginning of a word, when this word is closely associated with a preceding one which ends in a vowel, thus ܡܳܠ ܕ݂ܐܷܬ݂ܳܐ, John 16, 8; ܡܠܴܐ ܬ݂ܽܘܒ, John 16, 16; ܐܱܒ݂ܳܐ ܒ݂ܺܝ ܘܷܐܢܳܐ ܒ݂ܶܗ John 10, 38 (Bernstein) &c. The slightest pause, however, interrupts the softening. Similarly, two closely-associated words, of which the first ends in the same consonant as that with which the second begins, or a consonant like it, are so pronounced together that a doubling appears, which is indicated by the Q. of both of them: ܡܱܣܱܒ݁ ܒܱ݁ܐܦܷ̈ܐ massabbappē (instead of ܡܣܒ݂ ܒ݁ܐܦ̈ܐ "playing the hypocrite"; ܢܳܣܷܒ݁ ܒܱ݁ܐܦܷ̈ܐ "hypocrite"; ܒܷܬ݁ ܕ݁ܝܘܿܬ݂ܐ "ink-bottle".

Greek words. § 25. According to the prescriptions of the Schools, Greek words are not to be subjected to the rules for softening and hardening. Thus ܕܦ݁ܪܨܘܿܦܴ݁ܐ de̊πarṣōπā (πρόσωπον); ܡܶܢ ܦ݂ܶܝܠܻܝܦ݁ܘܿܣ "from Philippos", &c. (where ܦ݁ is the Greek π, § 15). ܒ݂ is made the equivalent of the Greek β, ܕ that of δ, ܟ݂ of χ ܦ݂ of φ; ܓ݂ generally that of γ. Thus for instance ܟ݂ܠܱܡܽܘܣ χλαμύς, ܬ݂ܐܘܪܝܐ θεωρία, ܒ݂ܣܝܠܠܝܘܣ Βασίλειος, ܓ݂ܪܱܡܱܜܺܝܩܹܐ γραμματική, &c. ξ has to be ܟ݁ܣ, e. g. ܐܷܟ݁ܣܘܿܪܻܝܱܐ ἐξορία; yet ܟ݂ܣ appears frequently, e. g. East-Syrian ܜܲܟ݂ܣܵܐ τάξις (West-Syrian ܜܶܟ݁ܣܴܐ). Generally speaking we find here too,—especially in words early introduced,—transformations, of a genuine Syrian type, e. g. ܐܷܣܟܹ݁ܡܳܐ σχῆμα, ܐܱܦ݂ܬ݁ܪܴܐ φθορά, ܣܽܘܒ݁ܠܴܐ συμβολή, &c.

Other foreign words too, in individual cases, vary from the rules, as regards R. and Q., e. g. ܦܷܬ݂ܓ݂ܳܡܳܐ "word" (Persian), where one would expect a hard ܓ.

DENTALS AND SIBILANTS.

Dentals and Sibilants. § 26. A. The ܬ of the Reflexive changes place, according to a common Semitic fashion, with the sibilant immediately following it (as first radical), and is altered into ܜ with ܨ, and into ܕ with ܙ, thus ܐܷܣܬ݁ܒܹܰܪ (for ܐܷܬ݂ܣܒ݂ܰܪ) "was thought", from ܣܒܱܪ "thought"; ܐܷܫܬܒ݂ܺܝ "was taken prisoner", from ܫܒ݂ܳܐ; ܐܷܨܜܠܷܒ݂ "was crucified", from ܨܠܱܒ݂; ܐܷܙܕܰܟܻ݁ܝ "was justified" from ܙܰܟܻ݁ܝ.

B. This ܬ is assimilated to a following ܜ and ܬ, becoming hard in the process: ܐܷܬ݁ܜܰܫܷܐ (pronounce eṭṭaššē) "was concealed"; ܐܷܬ݁ܬܱ݁ܒܱܪ (written also ܐܷܬ݂ܬܱ݁ܒܱܪ, ܐܷܬܬܱ݁ܒܱܪ) ettabbar "was broken in pieces"; so too, before a ܕ݁ furnished with a full vowel, e. g. ܢܷܬܕ݁ܰܟ݂ܪܴܟ neddakhrākh "remembers thee". A ܕ without a full vowel, on the other hand, here falls away in pronunciation, after the ܬ that has likewise become hard: ܐܷܬܕ݁ܟ݂ܰܪ ette̊khar "remembered"[7]. A like assimilation takes place, where an initial ܕ or ܬ without a full vowel is pressed by a foregoing prefix upon a following ܕ, ܬ or ܜ. The ܕ or ܬ is then written hard; ܘܱܕ݁ܕ݁ܡܷܐ "and who is like", ܠܱܕ݁ܕ݂ܰܩܕܩܺܝܢ "to persons or things, however small"; ܘܱܬ݁ܕ݂ܽܘܨ "and thou dost skip"; ܘܱܕ݁ܬ݂ܳܐܶܟ and who abides"; ܘܱܬ݁ܬ݂ܽܘܒ "and repentest, ܘܱܬܳܜܰܫܷܐ "and hidest": and the pronunciation must have been waddāmē, wattūṣ, etc. An ܜ or ܕ falls away before the ܬ݁ of a suffix in cases like ܥܱܒܻ݁ܝܜܬܴ݁ܐ ʿabbītā (or ʿabbittā?; West-Syr. doubtless ʿabītō) "thick (f.)"; ܦܫܺܝܜܬܴ݁ܐ "simple (f.)"; ܫܳܜܬ݁ܘܿܢ "ye despised"; ܠܴܜܬ݁ "thou didst curse", ܐܱܫܠܷܜܬܴ݁ܝܗ̄ܝ "gavest him power"; ܥܹܕܬܴ݁ܐ "Church"; ܡܨܺܝܕܬܴ݁ܐ "net"; ܥܒ݁ܕܬܴ݁ܐ "work"; ܐܷܒ݂ܰܕܬܷ݁ܝܢ "ye (f.) perished"; ܦܱܩܷܕܬ݁ܟ݂ܘܿܢ "I commanded you"; ܥܟܱܕܬ݁ "didst", and many others. In just the same way a pair of ܬ's coalesce, in words like ܐܱܒ݂ܗܶܬ݁ܬ݁ avhet or avhetht "madest ashamed"; ܐܱܒ݂ܗܶܬ݁ܬܴ݁ܢ "madest us ashamed" &c. The marking with R. and Q. varies; in effect, in all these cases only hard ܬ remains. For ܚܕܰܬܬܴ݁ܐ ḥe̊δattā "nova", one writes ܚܕܰܬܴ݁ܐ straight away, and ܚܕܽܘܬܴ݁ܐ "bride" for ܚܕܽܘܬ݂ܬܴ݁ܐ.

Radical ܕ falls away before ܬ in ܚܰܕ݂ܬ݂ܳܐ, ܚܱܿܕ݂ܬ݂ܳܐ, ܚܰܕ݂ܬ݂ܽܘܬ݂ܳܐ: pronounce ḥathā &c., "novus" &c.

C. A final ܬ has early dropped off in the absolute state of Feminines: ā coming from ath, ū from ūth, ī from īth, e. g. ܜܳܒ݂ܳܐ "bona"; ܜܳܒ݂ܽܘ "bonitas"; ܬܱܘܕ݁ܺܝ "confession"; in their construct state the ܬ remains: ܜܳܒ݂ܰܬ݂, ܜܳܒ݂ܽܘܬ݂, ܬܱܘܕܺܝܬ݂; and so also in the singular cases of ܘܠܴܢܺܝܬ݂ "a certain (f.)", and in many adverbs (§ 155).

D. Unusual is the assimilation found in ܓܶܦܴܐ "wing" from geδpā, as also the falling out in ܗܳܢܳܐ from hāδĕnā, and in other pronouns (§§ 67 Rem. 1; 68 Rem. 2).

LABIALS.

Labials. § 27. ܦ݂ܬ and ܒ݂ܬ are sometimes interchangeable. Thus ܙܶܒ݂ܬܴ݁ܐ frequently occurs for ܙܶܦ݂ܬܴ݁ܐ "pitch"; and occasionally on the other hand e. g. ܜܘܦܬܢܐ is found for ܜܽܘܒ݂ܬ݂ܳܢܳܐ "happy", and ܥܪܘܦܬܐ for ܥܪܽܘܒ݂ܬܴ݁ܐ "Friday". The East-Syrians have, from remote times, pronounced ܒ݂ quite like ܘ (w, u); av accordingly becomes au, uv, ū, e. g. ܫܘܼܒ݂ܚܵܐ šūḥā. They also pronounce ܦ݂ like ܘ, in cases where they leave it unusually soft and do not turn it into p (§ 23 A). Generally this transition is found in ܪܱܘܪ̈ܒ݂ܺܝܢ "magni", ܪܱܘܪ̈ܒ݂ܳܢܶܐ "magnates"; ܪܱܘܪܷܒ "made great", for רַבְּֿרְבִין,‎ רַבְֿרְבָנֵי,‎ רַבְֿרֵב‎ (§ 146). Compare ܫܱܘܫܡܳܢܳܐ "an ant" from ܫܱܡܫܡܳܢܳܐ (§ 31).

Liquids.

LIQUIDS.

n. § 28. N, as first radical, is almost always assimilated to the consonant immediately following it: ܐܱܦܷ݁ܩ "brought out", from anpeq; ܢܷܦ݁ܘܿܩ "goes out", from nenpoq; ܡܱܚܶܬ "brings down", from manḥeth; ܬܷܨܘܿܒ "plantest", from tenṣov, &c. Exception is made when ܗ follows: ܢܷܢܗܰܡ "roars"; ܢܷܢܗܰܪ "grows clear"; ܡܱܢܗܰܪ "lights", &c. (yet ܢܷܗܰܙ "thrusts" from nenhaz), and in other very rare cases (§ 173 A).

As second radical, n is assimilated in some nouns: ܥܷܩܳܐ "necklace"; ܠܷܬܴ݁ܐ "oppression"; ܐܱܦܷ݁̈ܐ "face"; ܓܰܒܴ݁ܐ "side"; ܐܱܦ݁ܬ݂ܳܐ "occasion"; ܫܷܕ݁ܬ݂ܳܐ "foundation", from ʿenqā &c.,—as against ܟܷܢܫܳܐ "congregation"; ܕܽܘܢܒܴ݁ܐ, ܕܽܘܢܒ݁ܬ݂ܳܐ "tail", which originally must have had a short vowel after the n, &c.

Farther, n loses its sound in many cases before ܬܴ݁ܐ of the feminine ending: ܓܦ݂ܶܬܴܐ ge̊fettā from ge̊fentā "vine"; ܓܒ݂ܶܬܴ݁ܐ "cheese"; ܠܒ݂ܶܬܴ݁ܐ "brick"; ܠܓܶܬܴ݁ܐ "a field-measure"; ܬܹܐܬܴ݁ܐ, ܬܹܬܴ݁ܐ[8] "fig"; and with n still written, in ܡܕܺܝܢ̄ܬܴ݁ܐ "town"; ܣܦ݂ܺܝܢ̄ܬܴ݁ܐ "ship"; ܙܒܱܢ̄ܬܴ݁ܐ "a time"; ܫܱܢ̄ܬܴ݁ܐ "year"; and in ܠܒ݂ܘܼܢ̄ܬܴ݁ܐ "incense", the n of which is still pronounced by others.

In ܓܰܢ̄ܒܴ݁ܪܴܐ gabbārā "hero", the nasal which serves as compensation for the doubling has been stroked out later.

On the dropping off of the n in the Imperative v. § 171 C, and in certain substantives, § 105.

l. § 29. L falls away when next to another l, in ܡܱܡܠ̄ܠܴܐ mamlā "speech", written also in fact ܡܡܠܐ; and in ܡܱܜܠ̄ܠܐ maṭlā "covering". Thus most Syrians say ܩܘܿܒܠܠܐ qovlā "countenance" (others qovelā).

It farther falls away in many forms which come from ܐܷܙܰܠ "to go" (v. § 183), as also in forms from ܣܠܷܩ (v. same section).

r. § 30. R falls out in ܒܱܪ̄ܬ݂ "daughter", construct state—(but not in the emphatic state ܒܱܪܬ݂ܳܐ).

Unusual Abbreviations with Liquids. § 31. We have unusual abbreviations in several nouns which are formed from the doubling of a short root ending in r, l, n, m: thus ܫܹܫܱܠܬܴ݁ܐ, ܫܹܝܫܱܠܬܴ݁ܐ "chain", from šelšaltā (cf. ܫܽܩ̈ܫܠܷܐ "tape-worms"); ܓܺܝܓ݂ܠܴܐ "wheel"; ܓܰܓܱ݁ܪܬܴ݁ܐ "throat" from gargartā; ܩܹܩܢܳܐ, ܩܹܝܩܢܳܐ "plough" from qenqe̊nā; ܫܱܘܫܡܳܢܳܐ[9] "an ant", probably from ܫܱܡܫܡܳܢܳܐ, and one or two others.

n becoming l in foreign words. § 31b. n beginning a word becomes l in several foreign words, like ܠܘܡܐ, along with ܢܘܡܐ, from νοῦμμος, nummus; ܠܱܡܜܳܐ with ܢܱܡܜܳܐ, from the Persian namat "carpet".

Gutturals.

GUTTURALS.

Falling away of initial ܐ. § 32. ܐ for the most part loses in Syriac its consonantal sound. As an initial sound it falls away along with its vowel in many words to which it belongs: ܐ̄ܢܳܫ or ܢܳܫ, ܐ̄ܢܳܫܳܐ, ܐ̄ܢܳܫܦܺܝܢ, ܢܳܫܺܝܢ "man", "men", &c.; ܐܺܚܪܹܝܢ or ܚܪܹܝܢ, ܐ̄ܚܪܹܬ݂ܳܐ, &c. "another"; ܐ̄ܚܪܴܝܴܐ or ܚܪܴܝܴܐ "last", ܐ̄ܚܪܴܝܬܷ݁ܗ "his last", &c.; ܐ̄ܚܝܳܢܳܐ "related"; ܐ̄ܢܳܐ or ܢܳܐ in certain cases for ܐܷܢܳܐ "I". Even in writing, this ܐ is without exception wanting in ܚܰܪܬ݂ܳܐ "end"; ܚܰܕ, ܚܕܳܐ "one" (m. and f.); ܚܳܬ݂ܳܐ "sister"; ܚܶܬܴ݁ܐ "pocket" (bag), and "bearing beam" (rafter) (v. אחד); ܬܴܐ, ܬܱܘ, &c. "come"; ܙܶܠ, ܙܶܠܝ, &c. "go"; ܘܱܙܳܐ, ܘܱܙܬܴܐ "goose", from אֱוַזָּא; ܕܶܝܢ (properly "then"[errata 2]) = אֱדַיִן.

Treatment of medial ܐ. § 33. A. As a medial, ܐ disappears completely according to the usual pronunciation, when it immediately follows a consonant or a mere sheva; and the vowel of the ܐ is transferred to the preceding consonant. Thus (a) ܡܱܜܶܐܒ݂ maṭev "makes good" for maṭʾev; ܢܶܫܱܐܠ "demands" for nešʾal; ܣܴܢܳܐܐ "hater"; ܜܰܡܱܐܬ݂ "unclean" f. (constr. st.) &c. (b) ܜܶܐܒ "was good" ṭev for ṭe̊ʾev; ܫܐܺܝܠܴܐ šīlā "demanded" (part.); ܦܷܐܐ "beautiful"; ܦܱ̈ܐܒ݂ܳܐ "beautiful" (pl.); ܟܳܐܬ݂ܳܐ "blaming", &c. So too after prefixes: ܕܰܐܒ݂ܳܐ "of the father", from ܐܱܒܳܐ + ܕ; ܠܾܐܘܡܳܢܳܐ "to the artificer" lummānā; ܘܷܐܟ݂ܰܠ "and ate"; ܒܱܐܝܕ݁ܳܐ "in what? (f.)" &c. In writing, such an ܐ is always left out in ܒܺܝܫ "bad", from בְאִישׁ, in ܡܱܠܷܦ݂, ܬܱܠܷܦ݂ "teaches", "teachest", &c. for מְאַלֵּף, &c.; farther, generally in the compound ܐܴܦܷܢ for ܐܴܦ݂ ܐܷܢ "although".

Although this falling away of the ܐ is very ancient, yet the East-Syrians frequently retain it as a consonant in such cases: thus e. g. they prefer to punctuate ܢܸܫܐܲܠ, ܒܐܵܬ݂ܵܐ, without pushing forward the vowel to the preceding consonant, as if it should still be read nešʾal, be̊ʾāthā; but all this without consistency.

B. Between two vowels ܐ receives with many Syrians (always?) the pronunciation y, e. g. ܐܴܐܱܪ ōyar "air" (West-Syr.). This pronunciation, which occasionally finds expression even in writing, e. g. ܨܰܝܱܐ for ܨܰܐܱܐ "defiled" (§ 172 A B), has however not been general.

In the end of a syllable ܐ always loses its consonantal value: ܫܷܐܠܷܬ݂ "I demanded", is in sound the same as ܫܷܠܷܬ݂; ܢܶܐܟ݂ܘܿܠ "eats" = ܢܷܟ݂ܘܿܠ; ܣܴܐܒ݂ܺܝܢ "are growing old" = ܣܴܒ݂ܺܝܢ, &c. Etymology alone can decide here, as in many other cases, whether ܐ is a mere vowel-letter or an original guttural (Arabic Hemza). Such an ܐ is now no longer written in cases like ܣܱܓܻ݁ܝ from saggīʾ (cf. ܣܴܓܻ݁ܝܐܴܐ, ܣܱܓܻ݁ܝܐܺܝܢ), &c.) "much". On the changes of vowels at the disappearance of such an ܐ v. § 53.

Auxiliary vowel of the ܐ. § 34. An ܐ, which in the beginning of the syllable ought to receive a vocal sheva,—according to the analogy of other consonants,—retains a full vowel instead; but in the middle of a word it gives up this vowel to the foregoing consonant (by § 33 A) and loses its own consonantal value. The vowel is ◌ܰ or ◌ܶ, and the latter even in many cases where it was originally a. Thus ܐܷܡܱܪ "spoke", compared with ܩܜܰܠ "killed" 3. s. (originally amar, qaṭal); ܐܱܡܺܝܪ "spoken", compared with ܩܜܺܝܠ "killed" (from qaṭīl); ܐܱܟ݂ܘܿܠ "eat", like ܩܜܘܿܠ "kill",—ܡܷܬ݂ܶܐܟ݂ܶܠ "is being eaten" (like ܡܶܬܩܜܶܠ "is being killed"); ܡܠܱܐܟ݂ܳܐ "angel" = מַלְאֲכָא; ܡܱܟ݂ܶܐܒ݂ܷܐ "afflicted" machevē (East-Syrian ܡܲܟܲܐܒܸܐ) &c. The Nestorians occasionally write in these cases ◌̱ (§ 17) e. g. ܡܸܬ݂ܐ̱ܟܸܠ, which is even improperly used for regular vowels, as in ܫܲܬܐ̱ܣܹ̈ܝܗܿ = ܫܷܬܐܷܣܷܝ̈ܗܿ (§ 45) "her foundations". An o (perhaps lengthened?) has been thus maintained in ܐܘܿܪ̈ܰܘܳܬ݂ܳܐ (Plural of ܐܘܿܪܝܳܐ "manger") from ŏrawāthā. Such an ܐ with a sheva disappears without leaving a trace in ܣܘܿܓܐܗܘܿܢ, ܣܘܿܓܗܘܿܢ "their multitude" from ܣܘܿܓܳܐܐ for soγʾā.

Orthographic Note on ܐ. § 35. Seeing that a radical ܐ frequently thus falls away in pronunciation, it is often left out also in writing, and that even in the oldest manuscripts, e. g. ܡܟܘܠܬܐ for ܡܷܐܟ݂ܽܘܠܬܴ݁ܐ "food"; ܢܟܘܠ for ܢܶܐܟ݂ܘܿܠ "eats"; ܦܬܐ for ܦܱܐܬ݂ܳܐ "face". On the other hand ܐ, even when a manifestly superfluous letter, is yet placed in words where it should not have appeared at all,—as in ܡܣܐܒ for ܡܱܣܱܒ "to take"; ܠܐܥܠܘܢ for ܠܷܥܥܽܘܢ "ye enter"; ܜܐܒܐ for ܜܶܒܴ݁ܐ "report"; ܩܐܝܡܝܢ for ܩܳܝܡܺܝܢ "stand" (pl.); ܬܘܗܐܝܐ for ܬܽܘܗܳܝܳܐ "delay"; ܕܐܘܘܢܐ, ܕܘܘܐܢܐ and even ܕܐܘܘܐܢܐ for ܕܵܘܘܿܢܵܐ or (West-S.) ܕܽܘܘܳܢܳܐ "pity", &c.; or it stands in the wrong place, like ܜܐܡܘܬܐ for ܜܰܡܽܐܘܬ݂ܳܐ "uncleanness"; ܫܐܘܠܐ for ܫܽܘܐܴܠܳܐ "question"; ܐܫܝܠܐ or ܫܺܐܝܠܴܐ "demanded" (part.) &c.; or it is doubled instead of being written once, as in ܢܒܝܐܐ for ܢܒ݂ܰܝܱܐ "comforts", and the like. The superfluous ܐ is a good deal in favour in certain causative forms, particularly in short ones, e. g. ܡܱܐܚܶܐ = ܡܱܚܶܐ "gives life"; ܢܰܐܗܰܪ "injures".

ܬܐ becoming ܬܬ. § 36. In certain cases a vowel-less ܬ݂, followed by an ܐ, blends with that letter into a hard ܬ݁ doubled and generally written ܬܬ (pointed ܬ݂ܬ݁, ܬ݁ܬ, ܬܬ̇, ܬܬ݁, which all express the same sound, § 26): in older days it was often signified by a single ܬ. Thus, regularly, in the reflexive of Aphel ܐܷܬܬܱ݁ܩܜܰܠ, ܐܷܬ݁ܬܱ݁ܩܜܰܠ for ethʾaqṭal; ܐܷܬܬ݁ܩܺܝܡ "was established" (ܐܬܩܝܡ) v. § 177 D &c. Thus, besides, in ܐܷܬܬ݁ܚܶܕ "was held" (ܐܬܚܕ) for ethʾe̊ḥeδ, and occasionally in similar forms (§ 174 C). A single ܬ is almost always written for ܬܬ, if another ܬ precedes by way of prefix, e. g. ܬܷܬ݁ܩܺܝܡ, ܬܷܬ݁ܚܶܕ, instead of ܬܬܬܩܝܡ, ܬܬܬܚܕ.

ܥ. § 37. Even before the orthography was elaborated, a ܥ‍ followed by another ܥ‍ in the same root became ܐ (ܐܶܠܥܴܐ "rib", from ܥܷܠܥܴܐ; ܐܱܥܦܴ݁ܐ "doubled", from ܥܱܥܦܴ݁ܐ, and many others):[10] In like manner, with the West-Syrians, a ܥ‍ coming immediately before ܗ becomes ܐ and is treated like it in every respect. Thus ܥܷܗܰܕ "remembered",—pronounce ܐܷܗܰܕ, from ܥܗܰܕ; ܥܽܘܗܕܳܢܳܐ "recollection",—pronounce ܐܾܘܗܕܳܢܳܐ; ܡܶܬ݂ܶܥܗܶܕ metheheδ for ܡܷܬ݂ܥܗܶܕ, &c. This change, which becomes noticeable even in the fourth century, and is occasionally indicated also in writing (ܐܗܪܝܢ, ܐܗܝܪ for ܥܴܗܪܻܝܢ, ܥܱܗܺܝܪ "to be in heat"), has however remained unknown to the East-Syrians.

ܗ. § 38. ܗ, which as an initial letter had, even in ancient times, often passed into ܐ (e. g. in ܐܷܢܘܿܢ secondary form of ܗܶܢܘܿܢ "they", and in the Aphel ܐܱܩܜܶܠ from haqṭel, &c.), falls away in pronunciation in many forms of the suffix of the 3rd sing. masc., e. g. ܡܱܠܟܱܘ̈ܗ̄ܝ malkau from malkauhī, "his kings"; ܒܢܳܝܗ̄ܝ "built it" (m.); ܢܷܩܜܠܻܝܘܗ̄ܝ "kills him". The personal pronoun—ܗܽܘ "he" or ܗܺܝ "she"—loses the ܗ, when it is enclitic, e. g. ܩܜܰܠܽ ܗ̄ܘܼ qe̊talū; ܠܷܗܽ ܗ̄ܘ or ܠܗܘ lēhū; ܡܳܢܳܐ ܗ̄ܝ mānāi from mānā hī; ܡܳܢܰܐ ܗ̄ܘܼ from mānā hū. In fact ܡܳܢܱܘ, ܗܳܢܱܘ, ܐܱܝܟܱܘ are often written for ܡܳܢܰܐ ܗ̄ܘ, ܗܳܢܰܐ ܗ̄ܘ, ܐܱܝܟܱܐ. So always ܠܱܘ "not", from ܠܴܐ ܗܽܘ. From ܗܘ ܗܘ, ܗܝ ܗܝ come ܗܽܘܝܽܘ, ܗܺܝܺܝ: but ܗܝ ܗܝ is occasionally written even yet, though we do not so often meet with ܗܘ ܗܘ.

The ܗ of ܗ̄ܘܳܐ "fruit", falls away when employed as an enclitic: ܩܜܰܠ ܗ̄ܘܳܐ, ܩܳܜܠܻܝܢ ܗ̄ܘܱܘ (§ 299), &c.

The ܗ of the very common verb ܝܗܒ "to give" falls away in the Perfect in all cases where it had a vowel; thus ܝܱܗ̄ܒ݂, ܝܱܗ̄ܒ݂ܬ݁, ܝܱܗ̄ܒ݂ܬ݁ܘܿܢ, ܝܱܗ̄ܒ݂ܬܷ݁ܗ, &c. The East-Syrians suppress the ܗ even in cases like ܝܸܗ̄ܒܹ݁ܬ݂, &c., and similarly in ܐܸܬ݂ܝܲܗ̄ܒܲܬ݂, ܒܸܗ̄ܠܹܬ݂, &c.

For ܝܺܗܽܘܕܳܐ "Judah", ܝܺܗܽܘܕܳܝܴܐ "a Jew", &c. (from יְהוּדָא, יְהוּדָיָא, &c.) one may say also ܝܽܗ̄ܘܕܳܐ, ܝܽܗ̄ܘܕܳܝܴܐ Yūδā, Yūδāyā. ܝܘܕ̈ܝܐ &c. are written even without ܗ.

Greek rh. § 39. In Greek words ܪܗ is often written to express the aspirated ῥ, e. g. ܪܗܘܡܐ Ῥώμη, ܦܪܪܗܣܝܐ (along with ܦܪܪܝܣܝܐ, ܦܐܪܪܝܣܝܐ and other forms of transcription) παῥῥησία, &c. This ܗ has no consonantal value, and only in mistake is it treated occasionally as a true consonant.

Vowel-Letters ܘ and ܝ. Usual changes.

THE VOWEL-LETTERS ܘ and ܝ

§ 40. A. W beginning a root becomes y in Syriac, as in Hebrew, when it is not protected by certain prefixes. Root WLD thus yields ܝܱܠܕ݁ܳܐ "child"; ܝܷܠܕ݁ܰܬ݂ "she bare"; but ܐܱܘܠܷܕ݂ "he begat"; ܡܱܘܠܴܕ݂ܳܐ "birth", &c. The initial w is however kept in ܘܱ, ܘ "and"; ܘܳܠܷܐ "it is becoming" (and so ܘܳܠܝܳܐ f.; ܘܳܠܝܳܐܺܝܬ "decently" &c.); ܘܱܥܕ݁ܳܐ "an appointment" (and thus ܘܱܥܷܕ݂ "to appoint"; ܐܷܬ݂ܘܱܥܱܕ "to agree upon"); ܘܱܪܻܝܕ݂ܳܐ "vein"; add the interjection ܘܳܝ "woe!", whence ܘܳܝܴܐ "the woe"; so too ܘܱܪܘܳܪܐ "bee-eater"; and ܘܱܜܐ "a kind of partridge", which two words evidently are meant to reproduce the natural calls of these birds. Other words beginning with ܘ like ܘܱܪܕܳܐ "rose" are foreign or uncertain.

B. ܘ and ܝ have both of them too much of the nature of vowels to be able to stand as true consonants in the end of a syllable; they always form in that case simple vowels or diphthongs, thus: ܫܽܘܘܕܳܝܴܐ "promise" (with ܫܱܘܕܺܝ šaudī "promised") šūdāyā, not šuvdāyā, for it was frequently even written with just one ܘ)[11]; ܠܱܘ lau "not", not lav (from lā-ū, lāhū § 38); ܘܪܱܘ (East-Syrian ܘܪܵܘ) "called" qe̊rau; ܓܰܠܺܝܘ "revealed" (3 pl.) gallīu (not gallīv)[12]; ܒܱܝܬܴܐ "house" baitā; ܩܳܝܡܺܝܢ "rise" qāimīn; ܐܘܿܪܗܳܝܬܴ݁ܐ "Edessena" Orhāitā, &c.

C. ܝ without a full vowel always becomes ī in the beginning of the syllable. In the beginning of a word ܐܝ is often written for it; thus ܝܺܬܷܒ, ܐܺܝܬܷܒ īthev "sat", from יְתֵב; ܝܺܕܱܥ, ܐܻܝܕܰܥ īδaʿ "knew", from יְדַע; ܝܺܕܰܥܬ݂ܳܐ, ܐܝܕܥܬܐ "knowledge"; ܝܺܪܱܚ, ܐܻܝܪܱܚ "month" (emphatic state ܝܱܪܚܳܐ); farther, ܘܺܝܬܷܒ or ܘܐܝܬܒ, ܕܝܕܥܬܐ or ܕܐܝܕܥܬܐ &c. In later times the ܐ is not so often written in such cases as it was in earlier days. But still the ܐ is always found in ܐܻܝܩܳܪܴܐ "honour", ܐܻܝܕܳܐ "hand", ܐܻܝܡܳܡܳܐ "day", and thus in ܒܺܐܝܕܳܐ, ܠܻܐܝܡܳܡܳܐ &c. On ܝܽܗ̄ܘܕܳܝܴܐ along with ܝܺܗܽܘܕܳܝܳܐ, and ܝܱܗ̄ܒ instead of ܝܺܗܰܒ v. § 38.

So too, within the word, ܢܷܬ݂ܺܝܗܶܒ "is given", from נֶתְיְהֵב;‎ ܚܰܕܺܝܗܘܿܢ "their breast", from הַדְיְהוֹן (ܚܰܕܝܴܐ); ܟܽܘܣܺܝܬ݂ܳܐ "cap", from כוּסְיְתָא; ܫܓ݂ܽܘܫܺܝܗܘܿܢ "their commotion", from ܫܓܽܘܫܝܳܐ, &c.

In a closed syllable ye or yi becomes ī in ܐܻܝܬ݂ "exists", and in the foreign names ܐܻܣܪܴܝܷܠ or ܝܺܣܪܴܝܷܠ "Israel"; ܐܻܫܡܱܥܷܝܠ "Ismael" (both with orthographic variants); ܐܻܝܙܰܪܥܷܝܠ (for יִזְרְעַאל); and ܐܻܝܣܚܳܩ. Quite exceptionally, other forms are found, v. § 175 A, Rem.

For ܝܷܫܽܘܥ "Jesus" the Nestorians say ܝܼܫܘܿܥ Īšōʿ.

D. In the middle of the word, ya becomes ī in the adverbial ending āīth, from and along with āyath (§ 155 A). ܘ, which appears as an initial letter without a full vowel only in ܘ "and" (A supra), is sometimes treated within a word just like ܝ. Thus from remote times there appear as alternative forms ܚܰܝܘܬ݂ܳܐ ḥaiwe̊thā and ܚܰܝܽܘܬ݂ܳܐ ḥayūthā "animal"; ܚܰܕܘܬ݂ܳܐ and ܚܰܕܽܘܬ݂ܳܐ "joy" (§§ 40 D; 101; 145 F)[13]: forms with ū in these cases have become more usual; while other forms,—for instance, ܠܷܐܽܘܬ݂ܳܐ along with ܠܷܐܘܬ݂ܳܐ (לֵאוְתָא) "weariness", ܕܢܲܪܘܼܚܘܼܢ (East-Syrian) along with ܕܢܱܪܘܚܽܘܢ "that they may have room"—occur only in isolated cases.

E. A ܝ after ā, and before another vowel, is pronounced by the East-Syrians like ܐ, thus ܚܳܝܷܐ "lives", ܐ̄ܚܪܴܝܱܬ "at last", like ḥāē, ḥe̊rāath, &c.[14] (thus the converse of § 33 B). Perhaps old modes of writing, like ܪܘܚܢܐܝܢ for ܪܽܘܚܳܢܳܝܺܝܢ "spiritual" (pl.), are founded upon this. If the vowel succeeding ܝ, after a or ā, is e or i, then the difference between the highly vocal y and ܐ is hardly perceptible. Whence come the interchangeable forms ܡܱܝܺܝܬ and ܡܱܐܺܝܬ "dead"; ܦܱܝܺܝܫ "remaining" and ܦܱܐܺܝܦ (§ 118); ܐܱܫܩܳܐܺܝܢܝ and ܐܱܫܩܳܝܺܢܝ, "give me to drink" (§ 196) &c.: Thus old MSS. have ܫܪܝܪܝܝܬ for ܫܱܪܻܝܪܴܐܻܝܬ "truly" (§ 155 A).

F. In the same way awu and aʾu are scarcely distinguishable by the ear. Accordingly we find, for example, ܪܡܘܘܢ or even ܪܡܐܘܘܢ for ܪܡܱܐܾܘܢ "they threw" (§ 176 E), ܡܚܘܘܗܝ or ܡܚܐܘܘܗܝ for ܡܚܰܐܾܘܗ̄ܝ "they struck him" (§ 192), &c. Similarly, ܡܠܘܘܐ as well as ܡܠܘܿܐܴܐ "matter".

G. ܝ serves in rare cases as a mark of a vowel and a consonant at one and the same time; e. g. in ܢܒܺܝܳܐ ne̊vīyā "prophet" (in which the conclusion must have a sound differing very little indeed from that in ܐܱܬܻ݁ܝܐܴܐ "come", &c); ܫܺܝܽܘܬܴܐ šīyūthā "form"; and in the before-mentioned ܐܱܫܩܳܝܺܢܝ ašqāyīn. Similarly ܩܘܪ̈ܝܝܢ for ܩܽܘܪ̈ܝܴܝܺܝܢ quryāyīn "rustici" (to avoid the triple ܝ).

H. The Greek ια, ιω, &c. are sometimes treated as monosyllables, sometimes as dissyllables, for instance: ܗܶܕܝܘܿܜܳܐ ἰδιώτης; ܐܱܟܣܶܢܝܳܐ ξενία, ܡܱܪܩܝܘܿܢ Μαρκίων, together with ܡܱܪܩܺܝܘܿܢ; ܕܝܱܬܹܝܩܹܐ διαθήκη (along with ܕܺܝܱܬܹܝܩܹܐ); ܦܪܷܜܘܿܪܻܝܘܿܢ πραιτόριον (and ܦܪܷܜܘܿܪܻܝܢ) &c.

ܘ and ܝ as representing the 2nd and 3rd radical. § 41. In Semitic inflection ā appears instead of a theoretical aya, or awa, e. g. qām(a) "stood", like qaṭal(a) "killed"; galāt (Syriac ge̊lāth) "she revealed", like qaṭalat: ī instead of awī, e. g. qīm "stood (part.)" for qawīm, &c.

But in these cases the question turns very little indeed upon actual sound-transitions. Of quite predominant importance here, are those ancient analogical modes of formation, which mount up to a time long before the separation of the several individual Semitic tongues.

2. Vowels.

2. VOWELS.

LONG AND SHORT VOWELS IN OPEN AND CLOSED SYLLABLES.

Long vowels. § 42. Long vowels in open syllables remain unshortened. Syriac however has closed syllables with long vowels, even in the middle of the word, e. g. ܩܳܡܬ݁ܘܿܢ "ye stood" (2. m. pl.), ܐܱܩܺܝܡܬ݁ܘܿܢ "ye raised", and later formations like ܒܪܻܝܟ݂ܬܴ݁ܐ (first from be̊rīkhe̊thā) "benedicta", ܝܴܬ݂ܒܻ݁ܝܢ "sit" (part.), ܐܱܥܻܝܪܬܷ݁ܗ "I awoke him", &c. The East-Syrians have a marked inclination to shorten long vowels in closed syllables, and accordingly they often write straight away ܥܲܠܡ̈ܝܼܢ "eternities", for ܥܵܠܡ̈ܝܼܢ, ܥܴܠܡܻ̈ܝܢ, &c., and so too in the final syllables of ܐܸܬܲܬ for ܐܸܬ݂ܵܬ݂ "she came", (ܐܷܬ݂ܳܬ݂), &c. On the other hand they incline to lengthen short vowels in an open syllable, if these are exceptionally retained, and thus, e. g., regularly write ܐܲܪܡܝܵܬܹܗ "she threw it (m.)" for ܐܱܪܡܝܱܬ݂ܶܗ.

Rem.—As they have ceased to notice that the ◌ܵ, which they perhaps write in ܟܵܬܒܝܼܢ but pronounce short, is a long vowel, they set down now and then ◌ܵ for short a, e. g. ܡܵܠܦܝܼܢ for ܡܲܠܦܝܼܢ, ܡܱܠܦܺܝܢ "they teach" (part.).

Short vowels. § 43. A. Short vowels in closed syllables remain; but in open syllables short vowels have, in Aramaic, at a very early stage passed mostly into sheva mobile. This occurrence is precisely what has given the language its characteristic stamp. Thus, for instance, ܩܜܰܠ qe̊ṭal from qaṭal "killed"; ܕܗܰܒ݂ from dahav (cf. ܕܰܗܒ݂ܳܐ) "gold"; ܡܱܡܠܟ݂ܺܝܢ from mamlīkhīn "are kings" (sing. ܡܱܡܠܷܟ), &c. Then in Syriac even the sheva mobile has often quite disappeared, as we are able in part to establish, even for very early times, through the relations of Rukkākhā and Quššāyā (§ 23 D): compare also the treatment of originally doubled consonants (§ 21 B).

B. A sharpened syllable does not count for an open one, even when the double-consonant is itself simplified (§ 21 A, B). Thus the short vowel remains, with resulting hardness, in ܪܱܒܻ݁ܝ (rabbī, West-Syrian rabī) "brought up"; ܪܷܒܻ݁ܝܬ݂ܳܐ "interest"; ܡܱܚܶܡ (maḥḥem) "heats"; ܩܽܘܜܳܠܴܐ (quṭṭālā) "murder"; and so even ܫܱܐܷܠ "asked"; ܫܽܘܐܴܠܳܐ "question" (for theoretical šaʾʾel, šuʾʾālā). Here and there the falling away of the doubling in the pronunciation is to be made up for by lengthening the vowel.

C. But still in certain cases a short vowel holds its ground even in an open syllable: thus with ܐ as the initial letter of a syllable (§ 34), e. g. ܡܱܠܱܐܟ݂ܳܐ for מַלְאֲכָא "angel"; in the secondary forms ܢܷܩܽܘܡ, ܢܱܣܺܝܡ for ܢܩܽܘܡ, ܢܣܺܝܡ "stands", "sets" (§ 177 C) ; in many later forms like ܩܜܰܠܽܘܢ, ܩ̈ܜܰܠܷܝܢ (§ 158 D); and in the forms of the Imperative with Object-suffixes like ܕܰܒܱܪܱܝܢܝ "lead me" (§ 190), &c. So also is it in forms like ܓܰܠܝܱܬ݂ܶܗ "she revealed it" (§ 152), a recent formation from ◌ܶܗ + ܓܰܠܝܱܬ. The Nestorians (always?) lengthen the a in such cases (§ 42).

D. Where there had been two open syllables with short vowels, one of these had of course to remain ; thus ܕܰܗܒ݂ܳܐ from dahavā "gold"; ܕܶܟ݂ܪܴܐ from dakharā "a male"; ܩܷܜܠܱܬ݂ from qaṭalath "she killed", &c.

E. So too, when the prefixes ܘ ܕ ܠ‍ ܒ‍ come before a vowel-less consonant, their vowel remains as an a[15], thus ܒܱܡܠܷܟ݂ from ܡܠܷܟ݂ + ܒ‍ "in a king"; ܠܱܓ݂ܒ݂ܳܪ "to a man"; ܕܰܩܜܰܠ "who killed"; ܘܱܢܣܱܒ݂ "and took". With the words mentioned in § 51, which may assume an ܐܷ as their commencement, the prefix ܒ is given as ܒܷ‍, and so with the other prefixes, thus ܒܷܫܜܳܪܴܐ "in the written bond"; ܠܷܫܬܴ݁ܐ "to the six", &c.

Thus too, a appears in the corresponding case, when several such prefixes come together at the beginning of a word: ܘܱܕ݂ܡܱܠܟܳܐ "et regis", from ܡܱܠܟܳܐ + ܕ + ܘ; ܘܱܠܕ݂ܰܒ݂ܩܷܜܠܴܐ "and to him that is involved in murder", from ܩܷܜܠܴܐ + ܒ‍ + ܕ + ܠ‍ + ܘ; ܘܱܒ݂ܕ݂ܰܩܜܰܠ from ܩܜܰܠ + ܕ + ܒ‍ + ܘ, &c. (but of course ܠܕ݂ܰܒ݂ܩܷܜܠܴܐ, ܒܕ݂ܰܘܩܜܰܠ, &c.).

If the second consonant of such a word is an ܐ, then the prefix usually takes the vowel: ܘܱܡܳܐܐ "and a hundred" wamā from wamʾā = מְאָא‎ + ו;‎ ܕܰܠܻܐܝ "who wearied" dalī from dalʾī; ܘܱܐܣܱܐܢ "and put on thy shoes" wasan from wasʾan, &c. And yet, along with these are also found, though ignoring the ܐ, forms like ܘܫܷܐܒ "and demanded" we̊šel = we̊ + šʾel (along with ܘܱܫܷܐܠ); thus, in particular, we most frequently have ܘܡܱܐܣܷܐ, ܘܢܰܐܣܷܐ, ܠܡܱܐܣܴܝܽܘ, and other forms from ܐܱܣܺܝ "to heal".

When two such prefixes stand before initial ܐ, the ܐ is generally neglected, e. g. ܘܒܱܐܝܢܳܐ "and in whom or what?", from ܐܱܝܢܳܐ + ܒ‍ + ܘ; ܘܠܷܐܡܳܟ "and to thy mother"; ܠܕ݂ܶܐܬ݁ܕ݁ܰܟ݂ܪܱܢ "to him who remembered us"; ܘܠܱܐܠܴܗܳܐ "et Deo", &c.—More rarely with ◌ܰ: ܠܱܕ݂ܰܐܗܡܺܝ "to him who neglected", from ܐܱܗܡܺܝ + ܕ + ܠ‍; ܕܰܒܽܐܘܪܚܳܐ "he who is on the way", &c. The same fluctuation is found with ܐܻܝ, ܝܻ, from ye: ܕܒ݂ܺܐܝܕ݂ܳܐ "who or what is in hand"; ܘܼܒ݂ܺܐܝ̈ܕ݂ܰܝܟ݁ܘܿܢ, with ܕܰܒ݂ܺܐܝ̈ܕ݂ܰܘܗ̄ܝ, ܕܿܠܝܼܫܘܿܥ (East-Syrian § 40 C); ܕܰܒܺܝܕܰܥܬܷܗ, &c.

Rem. The old poets express themselves in all these cases either with or without the a according to the requirements of the verse.

An ܝ, originating according to § 40 C, yields with such a prefix the forms ܠܻܝ, ܒܺܝ, &c., e. g. ܘܺܝܕ݂ܰܥ or ܘܐܻܝܕܰܥ "and knew", from ܝܺܕܰܥ, ܐܻܝܕܰܥ (= יְדַע) + ܘ.

Rem. The Nestorians oddly give the vowel a to the prefixes before ܝܗܘܕܐ, ܝܗܘܕܝܐ, &c., "Judah, Jew", thus ܠܲܝܗ̈ܘܼܕܵܝܹܐ, ܒܲܝܗܘܼܕ, ܘܐܲܝܗܘܼܕܵܐ, &c.

Some of the most important vowel-changes. ā.

SOME OF THE MOST IMPORTANT VOWEL-CHANGES.

§ 44. The ā is retained with the East-Syrians, but has become ō with the West-Syrians. The former also set down ◌ܵ for the most part to represent the Greek α, particularly in an open syllable,—for which the West-Syrians prefer to keep ◌ܰ.

Before n the transition from ā to ō is partly found even earlier; thus in the sporadically occurring ܬܡܘܢ, ܬܡܘܢܐ, ܒܣܡ̈ܘܢܐ, ܟܦܣܘܢܝܬܐ, &c., for ܐܱܡܳܢ "there", ܬܡܳܢܶܐ "eight", ܒܷܣ̈ܡܳܢܶܐ "spices", ܟܷܦܣܴܢܺܝܬܴܐ "menstruans"; in ܐܘܦ = ܐܴܦ "also"[16]; still more usual are ܢܷܣܝܘܿܢܳܐ "temptation" (from ܢܷܣܝܳܢܳܐ though somewhat different in signification ["test or trial" 2 Cor. 2. 9]); ܓܶܠܝܘܿܢܳܐ as well as ܓܶܠܝܳܢܳܐ "revelation"; ܝܱܪ̈ܩܘܿܢܶܐ "vegetables" &c. (§ 74).

a. § 45. a has frequently become e, e. g. ܩܷܜܠܱܬ "she killed", from qaṭalath (cf. ܩܱܜܠܷܗ "he killed him"); ܒܷܣܪܴܐ "flesh", from basarā, &c. Here and there the vocalisation fluctuates between a and e: the East-Syrians especially give preference, upon occasion, to the former; e. g. in ܐܲܠܲܨ alaṣ for ܐܷܠܱܨ "afflicted" (§ 174 A); ܫܲܬ݂ܐܸܣܬ݂ܵܐ for ܫܷܬ݂ܣܬ݂ܳܐ "foundation"; ܦܲܥܪܵܐ for ܦܷܥܪܴܐ "cavern"; ܪܲܗܜܵܐ for ܪܷܗܜܳܐ "course, run"; and in several others that have a guttural for the middle letter.

A š, immediately followed by another consonant, sometimes occasions e instead of a: ܐܷܫܟܱ݁ܚ, ܡܶܫܟܱ݁ܚ instead of aškaḥ, maškaḥ "find" (§ 164); ܡܶܫܬܻܝܬ݂ܳܐ "texture", contrasted with ܡܱܪܕܺܝܬ݂ܳܐ "course"; ܡܷܫܬܾ݁ܘܬ݂ܳܐ "feast" (but ܡܱܫܬܝܴܐ the same) overagainst ܡܱܪܕܽܘܬ݂ܐ "chastisement"; ܬܷܫܘܺܝܬ݂ܳܐ "bed", ܬܷܫܡܶܫܬܴ݁ܐ "service", contrasted with ܬܱܟ݂ܣܺܝܬ݂ܳܐ "covering", ܬܱܟ݂ܫܷܦ݂ܬܴ݁ܐ "petition" (but ܬܱܫܥܻܝܬܴܐ "narration")[17]: notice farther ܐܷܫܬܴ݁ܐ, ܐܷܫܬܻ݁ܝ, ܐܷܫܜܳܪܴܐ (§ 51). Similarly s in ܒ݂ܣܬܱ݁ܪ "behind", from ܣܬܱܪ + ܒ‍, where according to other analogies ba was to be expected.

ē. § 46. Within the word an ē has sometimes been produced through the quiescing of a consonantal ܐ, as in ܒܹܐܪܴܐ "well"; ܪܹܝܫܳܐ or ܪܹܫܳܐ "head"; ܢܹܐܦܴܐ "stone", ܟܹܐܢܳܐ, ܟܹܐܝܢ, or ܟܹܝܢ "right, just" (§ 98 C). In an open syllable ē is, without regard for etymology, expressed freely by ܐ, or even not expressed at all (and in the same way the Greek αι and ε are dealt with: thus even ܩܹܐܪܣܴܐ (qērsā = καιρός), while in a closed syllable ܝ (or even ܐܝ) is set down by preference: In later times ܝ is more prevalent; e. g. the old form ܢܦܐܫܐ, becomes later ܢܦܹܝܫܳܐ ne̊fēshā "refreshment, recovery"; and ܠܡܹܐܢܐ λιμένα "harbour" takes later the form ܠܡܺܝܢܳܐ, &c.

This ē became to a large extent ī with the West-Syrians: They said ܢܼܐܡܱܪ nīmar "says", ܐ̄ܚܪܹܢܳܐ ḥe̊rīnō "alius", ܪܝܫܳܐ rīšō, ܟܹܐܦܴܐ kīfō, ܟܹܐܝܢ kīn, &c. Yet they keep the ◌ܶ in ܢܶܐܟ݂ܘܿܠ "eats", ܡܶܐܟ݂ܽܘܠܬܴ݁ܐ "food", ܓܶܐܪܴܐ "arrow", &c.; and there are found still in isolated cases ܢܦܷܐܫܳܐ as well as ܢܦܺܐܫܳܐ, ܡܷܐܡܳܐ as well as ܡܺܐܡܳܐ (Inf.) "to swear", ܬܷܐܒ݂ܕ݁ܽܘܢ as well as ܬܻܐܒ݂ܕ݁ܽܘܢ "are lost" (2. m. pl.), &c. (§§ 174 A, 175 B). ܐ, ܐܝ—or the defective form of writing i,—are (even apart from etymology) in these cases almost invariably certain marks of an original ē. The style of writing of the East-Syrians separates ◌ܹ ē with tolerable consistency from ܝܼ ī[18].

In the end of a word the West-Syrian transition from ē to ī, except in ܢܹܐ (= Hebr. נָא) appears only in Greek words in η, e. g. ܕܺܝܱܬ݂ܝܩܺܐ or even ܕܺܝܱܬ݂ܝܩܺܝ διαθήκη for ܕܝܵܬܹܝܩܹܐ of the East-Syrians. Otherwise ◌ܶ remains here: ܓܳܠܷܐ "reveals", ܡܱ̈ܠܟܷܐ "kings", &c.

e. § 47. The short ◌ܸ seems to have been ĕ in the West, from ancient times; in the East it was pronounced sometimes as ĕ, sometimes as ĭ. This difference has no grammatical significance.

A short ĕ may often be lengthened in the concluding syllable through the (original) tone: thus ܕܳܚܶܠ "terrifies"; ܩܷܜܠܷܬ "I killed" (in which cases the second vowel is written by the East-Syrians with ◌ܵ) should perhaps be pronounced dāḥḗl, qeṭlḗth: It is the same principle with the monosyllabic ܫܷܠܝ (ܡܶܢ) "suddenly" and ܫܷܬ݂ "six", for which ܫܝܠ and ܫܝܬ are found in very old MSS. Yet this is not certain; and still less certain is it whether such a lengthening was generally practised. But beyond all doubt ܒܷܪܝ "my son" (§ 146) has a long ē.

ō, o. § 48. The ܘܿ (ō) with the West-Syrians at an early date coincided with ܘܼ (◌ܽܘ, ū). It has been retained only in the interjections ܐܘّ and ܐܘܿܝ "oi" (for which others say ܐܾܘܝ). Thus we have otherwise ܩܳܜܽܘܠܴܐ qōṭūlō for ܩܵܜܘܿܠܵܐ qāṭōlā "murder"; ܨܠܽܘܬ݂ܳܐ ṣe̊lūthō for ܨܠܘܿܬ݂ܳܐ ṣe̊lōthā "prayer", &c. Moreover such an East-Syrian ܘܿ appears not seldom to be only the result of toning down an original ū, especially in the neighbourhood of a guttural or an r, e. g. ܝܺܫܘܿܥ (§ 40 C), ܙܥܘܿܪܐ "small", ܚܪܘܿܪܐ "hole", ܫܡܘܿܥܬܐ "report", ܫܘܿܥܐ "rock", and many others: so too in the neighbourhood of an n, e. g. ܐܲܬ݁ܘܿܢܵܐ "oven", ܓܢܘܿܢܵܐ "tent". In many casesܘܿ may denote an o originally short, but lengthened by the tone; so perhaps in ܢܷܩܜܘܿܠ "kills", ܩܕܘܿܫ "sanctuary" (§ 103), &c. Still, there is as little certainty about tins as about the similar case in § 47.

The East-Syrians in particular distinguish also a short ܘܿ (o) from a short ܘܼ (u), but this distinction is of little importance. Here too a guttural or an r frequently seems to bring about the ܘܿ pronunciation, e. g.: ܬܸܫܒܘܿܚܬ݁ܐ "glory", ܐܘܿܪܝܵܐ "manger", &c.

It is curious that the West-Syrians have, besides the form ܟܽܠ "all", the form kol, which accordingly they have to write ܟܳܠ. Is it a lengthened kōl? So too ܟܳܠܷܗ, ܟܳܠܴܟ, &c.

While even with the East-Syrians the sound o began pretty early to pass into u, the tradition varies a good deal in the case of ܘܿ and ܘܼ; but with respect to cases of grammatical importance there is no doubt whatever.

Greek ο and ω are with the East-Syrians either retained,—and then they are written ◌ܳܘ, ◌ܳ, e. g. ܬ݂ܪܴܘܢܳܘܣ, ܬ݂ܪܴܢܳܘܣ θρόνος—, or they become u. There is a good deal of variation in the usage, e. g. ܦܺܝܠܻܝܦܴܘܣ and ܦܺܝܠܝܦܽܘܣ, ܗܺܝܓܡܳܘܢܳܐ and ܗܺܝܓܡܽܘܢܳܐ ἡγεμών &c.

With the East-Syrians ܘܿ corresponds to the Greek ο and ω, in so far as they keep from altering the words more decidedly.

As they cannot express an o without a vowel letter, they put ◌ܵ with defective-writing for the Greek ο, ω, and pronounce it ā, e. g. ܬܷܐܴܘܕܳܪܴܘܿܣ Theodāros for ܬܷܐܴܘܕܳܪܴܘܿܣ Θεόδωρος.

ai and au. § 49. A. The diphthongs ai and au remain very steady, particularly in the beginning of a word, although in dialects the pronunciation ē and ō occurred. Commonly, however, simplification of the diphthong prevails in a closed syllable. The West-Syrians farther proceed (according to § 46) to turn the ē occasionally into ī, and the ō always into ū (§ 48): thus, along with ܒܱܝܬܴ݁ܐ, ܒܷܝܬ݂ "house"; with ܚܰܝܠܴܐ, ܚܹܝܠ "strength"; with ܥܱܝܢܳܐ, ܥܹܝܢ "eye"; ܡܓܰܠܷܝܢ from me̊γallain, "they reveal"; ܬܪܷ̈ܝܢ from te̊rain, "two"; ܣܱܘܦܴ݁ܐ, ܣܘܿܦ݂, ܣܽܘܦ݂ "end", &c. So by analogy from ܠܥܹܝܢ (in oculo = coram) even in an open syllable ܠܥܹܝܢܘ̈ܗ̄ܝ, ܠܥܹܢܘ̈ܗ̄ܝ &c. coram eo; but only in the prepositional use; for example, otherwise, ܠܥܱܝ̈ܢܱܘܗ̄ܝ "to his eyes".

ܡܱܘܬ݁ܗܘܿܢ, ܡܱܘܬ݁ܝ "their, my death", ܥܱܝܢܟ݂ܘܿܢ "your eye", &c. form no exception, for in these cases it was only in the last development that the syllable became a closed one. Thus also is explained perhaps the retention of the ai before suffixes, in forms like ܡܱ̈ܠܟܱܝܟ݁ (from malkaikā), ܡܱ̈ܠܟܱ݁ܝܢ (from malkainā) "thy, our kings", and in verbal forms like ܓܠܱܝܬ݁ and ܓܠܱܝܢ (from ge̊laitā, ge̊lainā) "thou didst reveal", "we revealed". In ܠܱܝܬ݁ "is not", from ܠܴܐ ܐܻܝܬ݂, the diphthong is of more recent origin. On the other hand we have simplification in ܬܘܿܪܬܴ݁ܐ, ܬܽܘܪܬ݁ܐ "cow", from taure̊thā, and in East-Syrian ܠܹܠܝܵܐ, ܠܸܠܝܵܐ lēlyā, lelyā, West-Syrian ܠܠܝܳܐ, usually ܠܻܠܝܳܐ from laile̊yā "night". So too in ܒܹܝܥܬ݂ܳܐ, ܒܹܥܬܴܐ "egg" from baiʿe̊thā, pi. ܒܹ݁ܥܷܐ, ܒܹ̈ܝܥܷܐ.

B. The East-Syrians for the most part write ◌ܵܘ for ◌ܰܘ, and much more rarely ◌ܲܘ. So also in cases where the w is virtually doubled, as in ܚܵܘܝܼ = ܚܰܘܺܝ חַוִּי "pointed out"; ܬܩܵܘܹܐ תְקַוֵּא "thou remainest"; ܟܵܘܹ̈ܐ = ܟܱܘܷ̈ܐ "windows", &c. Thus too in ܡܵܪ̈ܵܘܵܬܵܐ = ܡܳܪ̈ܰܘܳܬܴܐ "Lords", and other plurals of that kind; farther in cases like ܐܲܫܩܵܐܘܼܗ̄ܝ = ܐܱܫܩܱܬܾܘܗ̄ܝ "give ye him to drink"; ܙܟ݂ܵܐܘܼܢܝ = ܙܟܱܐܾܘܢܝ "they overthrew me" (§ 192).

Sometimes on the other hand they write ◌ܲ‍ܝ for ◌ܳ‍ܝ, e. g. for "barefooted", and always in the Imperative ܩܜܘܿܠܲܝܗ̄ܝ = ܩܜܽܘܠܴܝܗ̄ܝ "kill him".

The West-Syrians also write an au produced by ā and u coming together,—with the vowel-sign ◌ܰ, e. g. ܡܱܠܟܱܐ ܗ̄ܘ malkau "is king", although the separate members are ܗܽܘ + ܡܱܠܟܳܐ. With them indeed ܡܱܠܟܳܐ ܗ̄ܘ would have the sound malkōu.

Loss of vowels.

LOSS OF VOWELS.

§ 50. A. Final vowels coming immediately after the original tone-syllable have all fallen away. This happened to ā even before the settlement of the orthography, thus ܠܱܢ from lánā "to us" ; ܐܱܢ̄ܬ from á(n)tā "thou" ; ܩܜܰܠܬ݁ from qe̊ṭáltā "hast killed", &c. (but ܡܱܠܟܳܐ malkā "king", &c). Other final vowels too have at quite an early date thus fallen away, without leaving a trace. On the other hand many vowels of this kind are still set down in consonantal character, although they had ceased to be pronounced even in the oldest literary epoch represented by documents (circa 200 A. D.)[19], and are ignored in punctuation. These are:—

(1) ū of the plural in the Perfect and Imperative after consonants: ܩܜܰܠܘ qe̊tal from qe̊tálū; ܩܱܜܶܠܘ, ܐܱܩܜܶܠܘ, ܩܜܘܿܠܘ; ܫܱܒܱ݁ܚܘ "they praised", &c. (but we have the full sound in ܓܠܱܘ ge̊lau, ܓܰܠܻܝܘ gallīu "revealed", &c).

(2) ī of the suffix of the 1st sing, after consonants, thus: ܡܱܠܟܝ malk "my king" from malkī; ܩܱܜܠܱܢܝ "killed me"; ܓܠܴܢܝ, ܓܰܠܝܱܢܝ "revealed me", &c. (but ܡܱܠ̈ܟܱܝ "my kings" ; and also the monosyllables ܒܺܝ "in me", ܠܻܝ "to me", in which no falling away was possible: So too ܟܘܿܠܝܼ, ܟܽܠܻܝ "I wholly", "the whole of me" ["my totality"]).

(3) ī of the suffix of the 3rd sing. m. ܗܝ with the noun: ܡܱܠ݁ܟܱܘܗ̄ܝ malkau from malkauhī "his kings", and with the Verb in cases like ܓܠܴܝܗ̄ܝ, ܢܷܩܜܠܻܝܘܗ̄ܝ, ܩܜܘܿܠܴܝܗ̄ܝ, ܢܷܓܠܷܝܘܗ̄ܝ, no doubt from ge̊lāihī &c.

(4) ī of the 2nd fem. sing. in ܐܱܢ̄ܬ݁ܝ at from a(n) "thou" (f.); ܡܱܠܟܷܟܝ malkḗkh from malkḗkhī (both with e?) ܡܱ̈ܠܟܱ݁ܝܟ݁ܝ; ܠܷܟ݂ܝ; ܩܜܰܠܬ݁ܝ; ܫܱܒܱܚܬ݁ܝ, &c.

(5) In the following special cases: in ܡܶܢ ܫܷܠܝ "from quiet" = "suddenly", absolute state of ܫܷܠܝܳܐ from šélī (like פֵּרִי); in ܐܷܡܱܬ݂ܝ "when?" from emmắthai; ܐܷܬ݂ܡܳܠܝ "yesterday" from ethmā́lē; and the derived word ܡܢܳܬܡܳܠܝ "the day before yesterday"; lastly in the much maimed form ܐܷܫܬ݁ܩܱܕܝ (or ܐܷܫܬ݁ܩܱܕ) "last year".

B. Even in very ancient MSS. the unpronounced ܝ's are often wanting: a similar ܘ is more rarely omitted. Conversely ܝ, which one was in the habit of so often writing,—apparently without cause,—was in some cases attached parasitically to words ending in a consonant; e. g. there occurs in old manuscripts ܐܠܗܝ for ܐܰܠܴܗ "God" (Construct State); ܐܒܝ for ܐܴܒ "August"; ܪܘܚܝ for ܪܽܘܚ "spirit". Occasionally it is employed as a diacritic mark of the 3rd sing. fem. of the Perf. e. g. ܩܜܠܬܝ for ܩܷܜܠܱܬ݂ "she killed". Such an employment of ܝ in the 3rd pl. fem. Perf. has gradually come into full use with the West-Syrians; ܩ̈ܜܰܠܝ "they (f.) killed", for the old ܩܜܠ retained by the East-Syrians (from original qe̊ṭálā, not qe̊ṭálī). The employment of ܝ in the 3rd sing. fem. Imperf.,—coming into view in rather late times,—prevails among the West-Syrians, though not quite so universally; ܬܷܩܜܽܘܠܝ, ܬܩܱܜܶܠܝ "she kills", &c, in order to distinguish it from the 2nd sing, masc, ܬܷܩܜܽܘܠ, ܬܩܱܜܶܠ "thou killest": the Nestorians are completely unacquainted with the ܝ in this usage.

New vowels and syllables. Vowel prefixed. (Alaf prosthetic).

NEW VOWELS AND SYLLABLES.

§ 51. An ܐ with a vowel is sometimes prefixed to an initial consonant which has not a full vowel. Thus ܐܷ in ܐܷܫܬܴ݁ܐ "six", ܐܷܫܬܻ݁ܝܢ "sixty", alongside of ܫܬܴ݁ܐ, ܫܬܻ݁ܝܢ; ܐܷܫܜܳܪܴܐ "a written bond" along with ܫܜܳܪܴܐ, and always ܐܷܫܬܻ݁ܝ "drank"; farther ܐܷܟܒܱܪ "already" sometimes for ܟܒܱܪ. Frequently so in Greek words with στ, σπ, like ܐܣܜܪܜܝܐ or ܣܜܪܜܝܐ στρατεία, ܐܣܦܝܪܐ and ܣܦܝܪܐ σπείρα, &c.

The prefix, pretty frequently met with in ancient MSS. before ܪ, is probably to be pronounced ܐܱ; e. g. ܐܪ̈ܚܝܡܐ for ܪ̈ܚܺܝܡܷܐ "Beloved"; ܐܪܕܝܕܐ for ܪܕܺܝܕܳܐ "upper garment"; ܐܪܩܝܥܐ for ܪܩܺܝܥܴܐ[20] "firmament"; ܐܪܥܐ for ܪܥܷܐ "contented", and many others. So too ܐܚܫܡܝܬܐ for ܚܫܳܡܺܝܬܴܐ "a meal" ; ܐܓܠܝܕܐ for ܓܠܻܝܕܳܐ "ice". In the frequently occurring ܐܾܘܪܩܱܥܬ݂ܳܐ the u of the rarer form ܪܽܘܩܥܬ݂ܐ, ܪܽܘܩܱܥ.ܹܬ݂ܳܐ is brought to the front. The early adopted Persian word rāzā ܐܪܙܐ, more rarely ܪܙܐ, ܪܐܙܐ "a secret" seems to have been pronounced with a vowel-prefix, which however is ignored in the pointing.

Auxiliary vowels. § 52. A. The poets sometimes insert an e before ܕ ܠ‍ ܒ‍ after a word ending in a consonant, e. g. ܐܝܬ ܠܗܘܢ "is to them" īth elhōn (with three syllables) = ܐܻܝܬ ܠܗܘܿܢ.

B. Essentially the same thing takes place frequently within the word. Especially when a consonant without a full vowel follows one that has no vowel, a short vowel is inserted often between the two to facilitate pronunciation. Thus ܡܱܕܶܢܚܳܐ = ܡܱܕ݂ܢܚܳܐ "sunrise"; ܕܶܚܶܠܬ݂ܐ = ܕܶܚܠܬ݂ܐ "fear"; ܬܷܫܱܒ݁ܩܽܘܢ = ܬܷܫܒ݁ܩܽܘܢ "you permit or remit"; ܬܷܕܰܚܠܻܝܢ = ܬܷܕ݂ܚܠܻܝܢ "thou fearest (f.)"; ܢܷܙܶܒ݁ܢܽܘܢ = ܢܷܙܒ݁ܢܽܘܢ "they buy"; also ܡܱܘܷܡܝܳܐ = ܡܱܘܡܝܳܐ "she swears"; ܙܰܘܷܥܬ݂ܳܐ = ܙܰܘܥܬ݂ܳܐ "quaking"; and ܫܸܐ̱ܠܬ݂ܵܐ (= ܫܷܐܷܠܬ݂ܳܐ v. infra C) "question". Particularly does this occur when one of the letters is a liquid or ܘ ܝ ܗ ܐ ܥ‍; on the other hand it is never found between sibilants and dentals. A marked amount of fluctuation however prevails in individual cases in the pronunciation of the various dialects and schools. With the old poets the longer forms, as indicated by the metre, are upon the whole rare; they abound in the vocalisation of the Bible, with both East- and West-Syrians.

C. The small stroke under the letter, called mehagyānā "the accentuator", serves as a sign of the fuller pronunciation particularly with the East-Syrians; the one above the letter, called marhe̊ṭānā "the hastener", as the sign of the shorter (§ 17). Yet often the full vowel is also written instead of the former, thus ܐܲܫ̱ܠܜܹܐ or ܐܱܫܷܠܜܶܐ = ܐܱܫܠܜܶܐ "I empowered".

The sign ◌̱ stands sometimes too in cases where the vowel which is supposed to be inserted is an original vowel, e. g. in ܩܸܩ̱ܠܬܵܐ = ܩܹܩܱܠܬܴ݁ܐ from qalqaltā. Sometimes it is not easy to say whether a vowel is original or inserted. Here and there such a vowel alters the original vocalisation more strongly; thus from ܥܱܩܪܒ݂ܳܐ "scorpion", has come the West-Syrian ܥܷܩܱܪܒ݂ܳܐ and then the East-Syrian ܥܩܲܪܒ݂ܳܐ.

The inserted vowel is mostly e, but often too it is a, especially before gutturals, and before q and r.

The relations of Rukkākhā and Quššāyā suffer no alteration through this insertion, as several of the foregoing examples show.

Influence of the consonants upon the vowels. Of ܐ.

INFLUENCE OF THE CONSONANTS UPON THE VOWELS.

§ 53. An ܐ originally a consonant and ending a syllable in the vowels. middle of a word becomes, in combination with a preceding a or i, an ē, which for the most part is farther developed with the West-Syrians into ī. Thus ܪܹܝܫܳܐ from רַאְשָׁא "head"; ܢܹܐܡܱܪ "says"; ܢܶܐܟ݂ܘܿܠ "eats"; ܐܹܡܱܪ "I say"; ܕܹܒ݂ܳܐ "wolf', from דַאְבָא, ܒܹܐܪܴܐ "a well" (also written ܒܪܐ § 46), and so forth.

On the other hand the ܐ becomes ā in ܥܴܢܳܐ "small cattle", through the influence of the neighbouring gutturals from עַאְנָא; ܥܴܥܢ̈ܳܬ݂ܳܐ "battlements" from עַאְעיתא; ܥܴܠܷܐ "a certain thorny shrub" from עַאְלא; and similarly ܚܰܢܳܐ "bosom" from חַאְנָא for original הַעְנָא.

In the end of the word we have ܢܹܐ from naʾ. In other cases ◌ܰܐ is retained here according to the analogy of corresponding forms ending in other gutturals, e. g. ܜܡܱܐ "unclean" (§ 100); ܜܰܡܱܐ "polluted"; ܒܱܝܱܐ "consoled" (§ 172), &c.

Of the other gutturals and of r. § 54. ܚ‍ ܗ ܥ and ܪ as final radicals, especially when they close the syllable, transform an ĕ into an ă; thus, ܢܷܕ݁ܰܥ "knows" (compared with ܢܷܬܷ݁ܒ "sits"); ܕܰܒܱܚ "sacrificed", compared with ܩܱܜܶܠ; ܢܒܱܗ "arose", for ne̊veh; ܢܕܰܒܱܪ "leads", for neδabber; ܐܱܘܕܰܥܢܱܢ "we made known"; ܢܒ݂ܰܗܬ݁ܘܿܢ "you arose"; ܦ݂ܪܱܚܬ݂ܳܐ "a bird"; ܕܰܒܱܪܬ݁ܘܿܢ "you led", &c. (§ 170).

In rare cases the transformation of an ܘܿ into a, before these final consonants, has been retained from very remote times, as for instance in ܢܷܦܬܱܚ "opens"; compare on the other hand ܢܷܕܒܘܿܚ "slaughters", &c. (§ 170). In certain cases they have the effect even of transforming a following e (or o?) into a (v. § 169).—On the exchange of a and e in words which have middle gutturals v. § 45.

On the shading off of an a into e through the influence of a sibilant, v. § 45; and of a u into o, effected by a guttural v. §§ 48, 49. In like manner the gutturals, as well as other consonants, particularly emphatic ones, must have brought about a special shading of the vowels in still other instances, without the writing giving much indication of such delicate turns.

3. Stronger alterations.

3. STRONGER ALTERATIONS.

§ 55. We find these, for instance, in the blending of Participles and Adjectives with the Subject-Pronouns: e.  g. ܩܳܜܠܻܝܬ݁ܘܿܢ (ܩܳܜܠܻܝܢ ܐܱܢ̄ܬ݁ܘܿܢ) from qāṭlīn a(n)tōn; ܩܳܜܠܻܝܢܱܢ from ܩܳܜܠܻܝܢ ܚܢܱܢ; ܒܪܻܝܟ݂ܰܬ݁ܝ "benedicta tu", from ܒܪܻܝܟ݂ܳܐ ܐܱܢ̄ܬ݁ܝ; ܩܳܜܠܱܬ݁ from ܩܳܜܶܠ ܐܱܢ̄ܬ݁ (§ 64 A), &c. Blendings with ܐܱܢ̄ܬ݁ appear in still other situations, e. g. ܕܰܗܒܳܐ ܐܢ̄ܬ݁ dahvat "thou art gold"; ܐܱܝܟܳܐ ܐ̄ܢ̄ܬ "ubi es?"; ܒܱܪ ܒܳܪܘܿܢܴܐ ܐ̄ܢ̄ܬ݁ bar bārōyat "thou art the son of the Creator"; ܕܚܰܝܷ̈ܐ ܐ̄ܢ̄ܬ݁ dēḥayyēt "vitae es", &c. Still in these cases the preservation of the separate portions is the more usual practice.

Amongst other instances we meet with extraordinary mutilations in the numerals of the second decade (§ 148 B); and farther in certain compounds (§ 141).

4. Tone.

4. TONE.

§ 56. The Nestorians now put the tone on the penult throughout, and that very distinctly. The Maronites[21], on the other hand, put the tone always, or almost always, on the last syllable, when it is a closed syllable, e. g. ܐܴܙܶܠ ōzél, ܩܷܜܠܱܬ qeṭlát, ܢܷܙܕܩܷܦ nezdqéf, ܝܱܘ̈ܡܺܝܢ yaumī́n, ܝܷܫܽܘܥ Ješū́ʿ, and so also in endings with a diphthong, e. g. ܐܷܬܱܘ etáu, ܬܱܠܡܺܝܕܰܘ̈ܗ̄ܝ talmīdáu, ܫܱܒܩܽܘܗ̄ܝ šabqū́i, ܐܷܒܢܷܝܘܗ̄ܝ ebnḗu. On the other hand they always, or nearly always, put the tone on the penult, when the word ends in a simple vowel: ܐܷܬܴܐ étō, ܢܺܐܬܷܐ nī́tē, ܨܳܒܷܐ ṣṓbē, ܢܷܗܘܷܐ néhwē, ܥܱܡܳܐ ʿámō, ܡܶܠܷܐ mélē, ܣܴܦܪ̈ܶܐ sófrē, ܗܳܢܳܐ hṓnō &c. Occasionally a secondary tone also becomes perceptible. At an earlier time the final syllable invariably had the principal accent.


  1. The proper name כזבי (Num. 25, 15) is written in Ceriani's Pesh. ܟܘܣܒܝ, where sb has the sound of zb. In Aphr. 111, 6, and Ephr. Nis. 71 v. 65 (in one Codex) it still stands ܟܘܙܒܝ.
  2. ܩܱܬܴ݁ܪܴܐ "stone" would form an exception, but this word is probably of foreign origin.
  3. And in that case, apparently, they always make it quiesce into u. Even the best Nestorian MSS. are, from these circumstances, of almost no value for an enquiry into R. and Q. of p. Besides even good MSS. and prints contain errors sometimes, as regards these 'points'.
  4. Contrary to the Hebrew כֹתְבִים, &c. A few exceptions, like ܦ݂ܠܓ݂ܺܝܢ 1 Cor. 9, 13, are cited.
  5. According to the best traditions.
  6. Even the best MSS. are not entirely free from error in their use of these points.—And in one or two cases, a distinction, founded upon R. and Q., has been established between words consisting of the same letters,—just through arbitrary pre- scription on the part of the Schools. Thus against all rules, they would have us say ܐܷܨܒ݂ܘܿܥ "I dye", but ܐܱܨܒܳܘܿܥ "I dip into"; farther ܐܴܚܕ݂ܺܝܢ "shut", but ܐܴܚܕ݁ܺܝܢ "hold", although these words are identical. The distinction, besides, between ܩܝܳܡܬ݁ܐ "resurrection" and ܩܝܳܡܬ݂ܐ was hardly known to the living speech. In addition to these examples there is a medley of cases resting upon the caprice of the Schools.
  7. Thus there are found in MSS. sometimes, forms like ܢܬܬܚܐ for ܢܷܬܕܚܶܐ "is pushed" and even ܡܬܬܝܚܘܬܐ for ܡܷܬ݁ܕ݁ܺܝܢܳܢܽܘܬ݂ܐ, ܡܷܬܬܪܺܝܢܳܢܽܘܬܴܐ "capability of being judged".
  8. Thus, with hard ܬ according to the best tradition. Probably the sing. of ܟܷܣ̈ܢܶܐ "corals" was pronounced as ܟܣܷܬܴ݁ܐ (Talmudic כסיתא).
  9. This vocalisation with au is much better supported than that with u (ܫܽܘܫܡܳܢܳܐ).
  10. Cf. ܢܳܢܥܴܐ "mentha" ['mint'] from נַעְנְעָא.
  11. Vice versâ,—because ܫܘܘ was pronounced like ܫܘ, the words pronounced šukōnō, šudōlō were in later times written ܫܘܘܟܢܐ, ܫܘܘܕܠܐ, where the doubled ܘ had no etymological foundation, since these words in their fundamental form are šukkānā, šuddālā, and belong to šakken "presented", and šaddel "enticed".
  12. The barbarous custom of pronouncing ו in the end of a syllable like a German w or indeed an f, instead of giving it a vowel sound (e. g. אביו aβīu, מלכיו me̊lākhāu), should be given up in Hebrew too.
  13. With the old poets these words are sometimes dissyllabic, sometimes trissyllabic. The Nestorians prefer the dissyllabic pronunciation of ܚܝܘܬܐ at least.
  14. Accordingly they like to put a small ܐ over such a ܝ.
  15. With ܘ and ܠ‍, a is the original vowel; perhaps ܒ‍ has just been adapted thereto by analogy, though originally it appears to have been bi and certainly analogy explains the treatment of ܕ, which is shortened from .
  16. ܦܘܪ̈ܣܝܐ "Persians" is probably an intentional defacement of the other and still more useful form ܦܴܪ̈ܣܴܝܷܐ: The hostile nation was denoted by a word which means "pudenda".
  17. ܡܶܫܢܳܐ "a pledge" is a borrowed word from the Assyrian, and accordingly does not belong to this class.
  18. Now-a-days the East-Syrians pronounce ◌ܹ,—both in cases where it corresponds to the ◌ܶ and in those where it corresponds to the ◌ܺ of the West-Syrians,—for the most part very like ī, and yet in another way than the pointed ܝܼ.
  19. Even the hymns of Bardesanes seem to neglect them, as regards the number of syllables.
  20. ܐܪܩܝܥܐ is measured as dissyllabic like ܪܩܝܥܐ in Moesinger's Monumenta Syriaca II, 86 v. 152 et passim, but ܐܪܕܝܕܐ, ܐܪܕܝܕܗ as trissyllabic in Jacob of Sarûg, Thamar v. 247, 251.
  21. I am indebted to my friend Guidi, following the communications made by P. Cardahi, for the data on the accentuation of the Maronites.

Errata:

  1. Original: Exceptions was amended to Exception: detail
  2. Original: "there" was amended to "then": detail