Some of you may have noticed that I use some weird symbols when I write Arabic words in Latin letters. This is to avoid ambiguity.
In English the digraph sh usually represents the sound found in the word “sheep.” So we might transliterate an Arabic word like شمس as shams. But what about أسهل? Would we transliterate it as ashal? In this case sh represents not one sound, but two sounds next to each other, /s/ and /h/.
To avoid ambiguities like this, we need to use one letter symbol per sound, and not use digraphs. In this particular case I use the elongated-S symbol <ʃ> to represent the sound in the word “sheep.”
My own style of transliterating Arabic is a mix of an academic style, such as Hans Wehr’s, and the International Phonetic Alphabet (IPA). I have generally tried to avoid using unsightly diacritic symbols, and so <ʃ> was preferred over <š> for the letter <ش>, and <ɣ> instead of <ḡ> for <غ>, etc. The only exception is the emphatic consonants, which I represent with a dot under the letter, e.g. <ṣ> for <ص>. I have also preferred the usage of the IPA symbols <ʔ> and <ʕ> for the letters <ء> and <ع> respectively, because the visual distinction between the left-facing hook <’> and the right-facing hook <‘> is not easy, especially on a computer monitor.
Below is a chart showing my transliteration scheme for consonants. On the right is the Arabic letter, in the middle is my transliteration, and on the left is a word with an equivalent English sound, if one is available.
For vowels I use a, i, u for the short vowels known in Arabic as fatħa, kasra, and ḍamma respectively. For their long vowel equivalents I use ā, ī, and ū. For the vowel-glide sequences, I use ay and aw for Classical Arabic and those dialects which preserve them as diphthongs. For dialects which monophthongize them, I use ē and ō respectively. For the epenthetic vowel I use the superscript letter ᵉ. For final tāʔ marbūṭa that has been fronted and raised, I use final -e or -et.