Vava ਵ

Vava is the thirty-fourth letter of the Gurmukhi painti and writes a semivowel that sits between the English v and w. Punjabi does not distinguish those two sounds, so a single letter serves for both and the realisation varies with the word, the speaker and the region without any of the variants being wrong. That is a recurring problem for romanisation, which has to pick one letter or the other and thereby suggests a distinction the script does not make; the same Punjabi word is commonly seen spelled both ways in English text. Vava is also the third and last of the letters that keep a subjoined form in Gurmukhi, with Haha and Rara, and that form appears regularly in the older orthography of the Guru Granth Sahib.

As a full letter it is frequent throughout Punjabi and throughout the scriptural vocabulary, so a reader meets both of its shapes early and often. On the page the full letter is a stem from the headline with a bowl to the left, and it is among the more distinctive outlines in the last third of the alphabet. The subjoined form of Vava behaves like the other two: it is written beneath the consonant it follows, leaving the headline unbroken, and it records that two consonants meet with no vowel between them. Because Gurmukhi kept only three such forms where its Indic relatives kept many, each of the three is common, and a reader of the Guru Granth Sahib meets all three regularly on a single page. Modern printed Punjabi uses them more sparingly, which is one of the differences between a scriptural page and an ordinary one, and a font built for general Punjabi may render the scriptural forms poorly or not at all without any visible error.

The pair of sounds Vava covers is a good place to state plainly what a romanised Punjabi can and cannot do. A romanisation is a mapping from one writing system to another, and it can only preserve distinctions that both systems make. Where Gurmukhi distinguishes retroflex from dental and the Latin alphabet does not, the mapping needs a diacritic or it loses eight letters into four. Where Punjabi makes no distinction between v and w and English makes one, the mapping has to invent a choice, and different transliteration schemes make it differently. Where Punjabi has tone and writes it with consonant letters, a romanisation that transcribes sounds rather than letters will produce something a Punjabi reader cannot map back to the spelling.

None of this is an argument against romanisation, which is necessary for indexes, catalogues, place names and any reference work with a mixed readership, but it is an argument for saying which scheme is in use and for keeping the original script alongside it wherever the exact letter matters. The atlas holds the Gurmukhi character for every letter as well as a romanised display name, so that nothing depends on the transliteration being reversible. The same principle applies to the display of the letters themselves. A romanised name is a label for a reader who does not know the script; the letter is the thing the label names, and only one of the two is authoritative.