The orthographic impact of intergenerational
phonological change in Indian Sindhi

By Arvind Iyengar, University of New England, Australia

“Studies have shown that spoken Sindhi, both in Pakistan and India, has been undergoing phonological changes from generation to generation, from at least the mid-20th century (Bughio 2001; Iyengar 2017). These changes involve both vocalic as well as consonantal transformations, resulting in the emergence of two different chronolects (Frellesvig 1996), or varieties spoken by the older and younger generations. Bughio (2001) terms these chronolects ‘Old’ and ‘New’, respectively.

The Old chronolect is most evident in the speech of Sindhi speakers aged 60 years or older, while the New chronolect is characteristic of the speech of those aged 40 years or younger.

These phonological changes in spoken Sindhi can have grapholinguistic implications, namely, implications for Sindhi orthography. In India, where a debate has raged on whether to write Sindhi in the Perso-Arabic or Devanagari script (Asani 2003; Daswani 1979; Iyengar 2017), the aforementioned changes in phonology may impact orthography in both scripts, leading to potential pedagogical issues.

Let’s take a closer look at these potential grapholinguistic implications in both scripts, starting with the Perso-Arabic script. This is an abjad or consonantary (Daniels 1996: 4), in that its basic graphematic units represent consonants. Certain vowel phonemes are represented by diacritics positioned above or below a base grapheme. For example, [məzᶦbut̪ᶷ] would be written ,مَضِبو ت where the three vowels [ə ɪ ʊ] are written using the diacritics , and , respectively. However, in everyday writing, these vowel diacritics are often omitted. The expectation is that fluent readers of Sindhi will mentally add these diacritics where needed. Thus, [məzᶦbut̪ᶷ] would typically be written مضبوت , sans diacritics, but a fluent reader of Sindhi would, in theory, read and comprehend it as if it were مَضِبو ت . Omitting optional diacritics results in less visual clutter and a neater appearance, but makes it harder for nonfluent speaker-readers of Sindhi to ‘fill in the gaps’.

In contrast, Devanagari is an abugida or alphasyllabary (Daniels & Bright 1996: xxxix). While its representation of vowels is similar to that of a consonantary, its representation of vowels is rather different. Devanagari it requires vowel diacritics to always be written. That is, unlike in Perso-Arabic, vowel diacritics can never be omitted in Devanagari. Thus, if [məzᶦbut̪ᶷ] were to transparently transcribed into Devanagari-script Sindhi, it would be written मज़ि बूतु – and always so, with none of the vowel diacritics considered optional or omissible. Compulsorily including all vowel diacritics in this manner results in less orthographic flexibility, but makes it easier for nonfluent speaker-readers of Sindhi to phonological decode any unknown words.

While Sindhi orthography in the Perso-Arabic script is relatively more standardised than that in the Devanagari script, given that the former has been in official use since 1853 (Boivin 2008), both continue to exhibit areas of variation (Iyengar 2018). Nevertheless, both orthographies are largely based on, and reflect, the Old variety chronolect of Sindhi. Consequently, with the emergence of a New chronolect, there is potential for a ‘mismatch’ between the Old-chronolect-based orthographies, and the New variety phonology. Even then, the mismatch may be more stark in one writing system than the other.

Let’s look at the implications of the intergenerational phonological change on Perso-Arabic Sindhi orthography. As seen in Table 1, the word pronounced [məzᶦbut̪ᶷ] in the Old chronolect is pronounced [məzbut̪] in the New chronolect. Orthographically, the diacritical Perso-Arabic spelling مَضِبوتُ is reflective of the Old pronunciation. Phonetically speaking, this spelling represents vowels that are no longer pronounced in the New chronolect. In fact, one might argue that a spelling that is more transparent and reflective of the New pronunciation [məzbut]̪ is مَضبو ت , with only the diacritic for [ə] written. Nevertheless, even if the vowel diacritics for [ɪ] and [ʊ] were to be omitted, it is evident that the graphematic modifications involved are minimal. Furthermore, were vowel diacritics to be altogether omitted—as is conventional in Perso-Arabic Sindhi orthography—we would end up with the same spelling mentioned earlier: مضبوت . From this, it emerges that, when faced with the nondiacritical form مضبو ت , an Old chronolect speaker-reader would mentally interpret the word as مَضِبوتُ [məzᶦbutᶷ̪], while a New chronolect speaker-reader would interpret the word as مَضبو ت [məzbut]̪.

On this basis, it could potentially be argued that the Perso-Arabic script makes a good candidate for an economical and compact pan-chronolectal Sindhi orthography.

In contrast, if one were to transparently represent the New chronolect pronunciations of words in Devanagari, the nature and structure of the script would require any graphematic modifications to necessarily be more conspicuous. So, while the Old chronolect pronunciation [məzᶦbut̪ᶷ] would be accurately reflected by the spelling मज़ि बूतु, the New chronolect pronunciation [məzbut̪] would require the spelling मज़्बूत् in order to be transparent. Even to a nonreader of Devanagari, it is evident that the graphematic differences between the two spellings is substantial.

Table 2 provides a summary of the correspondences between Old and New chronolect pronunciations, and their respective phonetically transparent spellings in the Perso-Arabic and Devanagari scripts.

Besides [məzᶦbut̪ᶷ ~ məzbut̪], also included in Table 2 for comparison purposes is the English loanword ‘card’.

As evinced by Table 2, transparently reflecting Old and New pronunciations in Perso-Arabic requires minimal graphematic changes. And when all diacritics are omitted per convention, the graphematic change in the orthography is nil. In contrast, the Devanagari spellings that reflect the Old and New pronunciations are graphematically and visually quite different. In short, Perso-Arabic allows for vowel diacritics to be omitted, which makes it difficult for nonfluent readers, but also makes it a potentially better candidate for a pan-chronolectal Sindhi orthography. In contrast, Devanagari does not allow for vowel diacritics to be omitted, which makes it easier for nonfluent readers, but also hinders the creation of a pan-chronolectal Sindhi orthography.”

Read the full article here : MIFS Newsletter  n°7-8 (July 2020)

---

---