Secondary Stress and Vowel Lengthening in Biblical Aramaic.

• Author: Aim, Emmanuel

1. INTRODUCTION

   Tiberian Biblical Aramaic suprasegmental features are indicated by two distinct means. The first consists of a set of accents that marks main stress and occasionally secondary stress. It also indicates the syntactic structure and the musical/prosodic phrasing of the verses. (1) The second is a diacritic called ga'ya which denotes inter alia secondary stress. (2) It is called major ga 'ya when indicating secondary stress on open syllables. Here the generalizations for main and secondary stresses are primarily based on the data and descriptions of Bauer and Leander(1927). (3)

   The fact that the Tiberian Masoretes accentuated Biblical Aramaic by using the system of notation employed for Biblical Hebrew leads some scholars to suppose that Aramaic stress rules reflect those of Hebrew. (4) Yet this is not completely true. While the rules governing the placement of the secondary stress in Biblical Aramaic are mostly identical to those of Biblical Hebrew, main stress placement of Biblical Aramaic differs significantly from that of Biblical Hebrew. (5)

   Thus, on one hand, it seems that Biblical Aramaic secondary stress should be treated as related to that of Biblical Hebrew. The Masoretes punctuated both parts of the Tiberian corpus, and it is probable that they pronounced them identically.

   On the other hand, it is also plausible that the common distribution of the secondary accent is not due to the same traditional way of intoning the texts or to the influence of one language on another, but to an independent parallel development.

   At any rate, before making assumptions and reaching more certain conclusions, it is a methodological prerequisite to examine Biblical Aramaic for its own sake (and not as mere variant of Biblical Hebrew). Although close to the phonological system of Biblical Hebrew, the phonological system of Biblical Aramaic is nevertheless different. Accordingly, the Tiberian signs denoting secondary stress should be first examined in the light of the structure of the language for which they are used. (6)

   To this day only two studies have offered an in-depth focus on the secondary stress in Tiberian Biblical Aramaic. The sole exhaustive description can be found in the authoritative grammar of Bauer and Leander (1927). While the authors provide a short survey of the possible historical reasons that have led to the current placement of the main stress, they do not determine the motivations for the various locations of the secondary stress. The synchronic analysis of Rothstein (1981) also investigates both main and secondary stresses. Yet, though Rothstein claims that her study is based on Bauer and Leander (1927), she ignores without justification some significant data. As a result, her analysis is biased. In the following sections, I shall reconsider this issue.

2. SECONDARY STRESS PLACEMENT

   The secondary stress regularly falls on an open syllable that is separated from the main stress by at least one syllable. (7) Thus, it is found regularly on 1) closed-syllable long vowels, e.g., (8)

   talattehon, 'the three of them' beltesassar, 'Belteshazzar' 'aparsakaye, 'Persians' unbazbeyatak, 'and your (ms) rewards' and 2) open-syllable long vowels, e.g., kateba, 'writing (fs)' (9) masroqita, 'the pipe' hakkimayya, 'the wise men' nebukadnessar, 'Nebuchadnezzar' It also falls on open-syllable short vowels, which appear in the following configurations: 1) before a schwa mobile; the two attested cases are the following:

   'amemayya, 'the peoples' (10) tiqepat, 'she grew arrogant' (11) 2) before a hatef, that is, an ultra-short vowel (which results from epenthesis or reduction), e.g.,

   sa'ata, 'the brief time' lemeheze, 'to see' loqobel, 'in front of 3) before an epenthetic short vowel (which takes the place of a hatef in closed syllable), e.g.,

   ta'abdun, 'you (mp) will do' hohorbat, 'she was laid waste' lehewyan, 'they (f) will become' 4) before a virtual geminate:

   /lahe"ala/ > lehe'ala, 'to come in' /me"ale/ > me'ale, 'the sunset' (12) /nahhiru/ > nahiru, 'illumination' /rahhiqin/ > rahiqin, 'keep aloof (mp)' (13) and 5) in the following unique case: (14) benaytah, I built her' Note that a word or a compound word can contain more than one secondary stress if there are sufficient target syllables available. In this case too, stressed syllables are separated from each other by at least one syllable. (15) For instance,

   lahaqamuteh, 'to set him' lahatabutak, 'to return to you (ms)' di-'aryawata, 'of the lions' di-biruselem, 'that in Jerusalem' By contrast, the secondary stress does not strike ultra-short vowels (schwa and hatef) and short vowels in closed syllable (exceptions to this generalization will be treated below). Accordingly, the following examples have no secondary stress:

   behilu, 'haste' habula, 'the crime' 'arkubbateh, 'his knees' 'ahasdarpenin, 'satraps' The fact that ultra-short vowels are not stressed is certainly not surprising since stress, from a cross-linguistic view, typically does not fall on reduced vowels. (16) On the other hand, the distribution between the accentuable CVVC, CVV, CV syllables, and unaccentuable CVC syllables is quite unexpected. Indeed, typological and theoretical studies on stress have long highlighted that quantity-sensitive stress systems always distinguish between heavy CVV syllables (as well as ultra-heavy CVVC) and light CV syllables. (17) Invariably, in certain contexts, stress will dock on heavy syllables whereas in the same contexts it will skip the light syllables. Since CVC syllables are treated as heavy or light under language-particular parameters, the distribution between stressed syllables and unstressed syllables are of two possible types (see Table 1). (18)

   In contrast, the Biblical Aramaic secondary stress pattern is apparently the following, where CV is counted as heavy (Table 2).

   Certainly, alternative attempts at rationalizing stress systems that do not fit the universal heavy accentuable CVV vs. light unaccentuable CV distinction can be advocated. Firstly, one can assume that the assignment of the secondary stress in Biblical Aramaic is not quantity sensitive. This is the perspective of the generative study of Rothstein (1981), which presents a set of rules to build a metrical tree that derives both main and secondary stresses. Yet these rules do not take into account the cases where the secondarily stressed CV syllables are followed by a schwa mobile (as in 'amemayya) or a virtual geminate (as in nahiru), or the special case of benaytah. Certainly, one could claim that in these cases the symbol ga 'ya is not used for indicating the secondary stress, but to alert the reader (as a kind of nota bene) of the unexpected occurrence of a short vowel in open syllable. However, Rothstein does not even discuss why she disregards these data. As a result, her analysis cannot be considered satisfactory. (19)

   Secondly, one can argue for a language-specific hierarchy heavy CVVC, CVV, CV > light CVC, CV to describe the secondary stress in Biblical Aramaic. For instance, language-specific heaviness and lightness without correlation to vowel length is proposed by Laks (1993: 34-41) for the stress system of Ossetic, a language without phonetic vowel length contrast.

   However, from a typological point of view, we can presume that there is no reason for Biblical Aramaic to be a language with odd stress properties. From a theoretical point of view, if Universal Grammar exists, we expect its principles to handle features of all languages, preferably without resorting to atypical language-specific parameters. For instance, Hayes (1980: 62) brings the Ossetic stress system into a more general pattern than Laks (1993) by analyzing the vowels that attract stress as phonologically long, and those that do not as phonologically short.

   Finally, it should be noted that the traditional point of view of Aramaists does not help in understanding the assignment of the secondary stress. Aramaists link vowel reduction with stress. (20) That is, counting from right to left, short vowels in open syllables are deleted if preceded by a CV syllable (e.g., /zimana/ > zimna 'the time') and reduced elsewhere (e.g., /malkat/ > malkata 'the queen', /katiba/ > kateba 'writing (fs)'). Data where CV syllables are secondarily stressed are understood accordingly: the short vowel is stressed, thus not reduced. However, this explanation sets aside the question of the placement of the stress: why does the secondary stress target these short vowels?

3. SECONDARY VOWEL LENGTHENING

   In this section, I shall argue that the analysis of vowel quantity I have used so far is not completely correct. That is, the problem of the accentuation of short vowels in open syllables lies not with stress theories or typological universale, but with the analysis of the length of these vowels. Indeed, to all intents and purposes, short vowels in open syllables function as long vowels with respect to their attraction of secondary stress. If we suppose they are in fact long, then their behavior is no different from that of the other long vowels.

   Evidence for length--independent of the secondary stress issue--is the resistance of Aramaic to short vowels in open syllables. To suppose that original short vowels in open syllables have undergone secondary lengthening would be consistent with Aramaic phonotactics. Indeed, since Biblical Aramaic does not tolerate short vowels in open syllables, a short vowel in this position tends not to remain as such. As commonly known, the regular process is to drop or reduce...