Inscriptional Pahlavi Alphabet Table
| Letter | Unicode name | Phones (Middle Persian) | Translit Iranian | Translit Aramaic |
|---|---|---|---|---|
| 𐭠 | Letter ALEPH | a, aː | ˀ | A |
| 𐭡 | Letter BETH | b, w | b | B |
| 𐭢 | Letter GIMEL | ɡ, j | g | G |
| 𐭣 | Letter DALETH | d, j | d | D |
| 𐭤 | Letter HE | h | h | – |
| 𐭥 | Letter WAW-AYIN-RESH | w, o(ː), u(ː) | w | W |
| 𐭦 | Letter ZAYIN | z | z | Z |
| 𐭧 | Letter HETH | h, x | h | Ḫ |
| 𐭨 | Letter TETH | t̰ | ṭ | Ṭ |
| 𐭩 | Letter YODH | j, e(ː), i(ː), d͡ʒ | y | Y |
| 𐭪 | Letter KAPH | k, ɡ | k | K |
| 𐭫 | Letter LAMEDH | l, r | l | L |
| 𐭬 | Letter MEM-QOPH | m | m | M |
| 𐭭 | Letter NUN | n | n | N |
| 𐭮 | Letter SAMEKH | s, h | s | S |
| 𐭯 | Letter PE | p, b, f | p | P |
| 𐭰 | Letter SADHE | t͡ʃ, d͡ʒ, z | ṣ | Ṣ |
| 𐭱 | Letter SHIN | ʃ | š | Š |
| 𐭲 | Letter TAW | t, d | t | T |
| 𐭸 | One | number | non-positional numeral 1 | – |
| 𐭹 | Two | number | non-positional numeral 2 | – |
| 𐭺 | Three | number | non-positional numeral 3 | – |
| 𐭻 | Four | number | non-positional numeral 4 | – |
| 𐭼 | Ten | number | non-positional numeral 10 | – |
| 𐭽 | Twenty | number | non-positional numeral 20 | – |
| 𐭾 | One Hundred | number | non-positional numeral 100 | – |
| 𐭿 | One Thousand | number | non-positional numeral 1000 | – |
Inscriptional Pahlavi is the stone-cut ancestor of the Pahlavi scripts: the earliest attested form, first seen on clay fragments dated to the reign of Mithridates I (r. 171–138 BC), then on Parthian coins, and finally on the rock faces of the Sasanian empire - kings and grandees, down to the powerful high priest Kartir, carving Middle Persian where everyone could read it. The classic monument is the c. 293 CE Paikuli tower inscription of king Narseh.
Its defining trick is the huzwārišn system: the text is written with Imperial Aramaic words that you read as Persian - Wikipedia’s definition requires a script to be “both Aramaic-derived” and to “feature huzwārišn” before it counts as Pahlavi at all. You see mlk’ (“king” in Aramaic) and say šahr. That is why this chart carries two transliteration columns: lowercase spells the Iranian sounds; capitals spell the Aramaic words underneath.
The block (U+10B60) holds 27 named codepoints: Wikipedia’s “19 non-joining letters written from right to left, and usually with spaces between words”, plus eight non-positional numerals; five codepoints are unassigned. Below, each letter row also shows which sounds the single glyph covers - and two names, WAW-AYIN-RESH and MEM-QOPH, warn you up front that one shape may answer to several letters.
How to use
- Read a word in two passes: first pass, decode the Iranian column left-to-right values (b, g, d...); second pass, look for a capital-letter spelling that matches an Aramaic word - if you find one, the written form is a logogram and the pronunciation you say aloud is the Persian word, not the Aramaic one. The Hebrew-looking spelling mlk’ reads šahr “city/king-land” in context.
- Treat the merged names as a sound warning: WAW-AYIN-RESH means one glyph serves w, y and r; MEM-QOPH means one glyph serves m and q. Context (and, in coins, small form differences) decides - Book Pahlavi later pushed this to 13 graphemes for 24 sounds, the ambiguity this inscriptional script only begins.
- Use the numeral rows as written: they are non-positional signs (1, 2, 3, 4, 10, 20, 100, 1000) summed additively, the same eight-sign set the Parthian sibling block uses for its dates - handy for reading regnal years on coins, where ŠNT plus digits gives the era year.
Frequently asked questions
Why does every letter have two transliterations?
Because the script is a bilingual system in one line: Aramaic words are used as heterograms (huzwārišn, “archaisms”) and read aloud as their Iranian equivalents. The lowercase Iranian column records what you say; the capital Aramaic column records what you see. A form “only qualifies as Pahlavi when it is both Aramaic-derived and features huzwārišn.”
What is the oldest surviving Pahlavi writing?
Wikipedia pins the earliest evidence to clay fragments from the reign of Mithridates I (r. 171–138 BC), with the Parthian coin legends next; the famous monumental phase is Sasanian - royal rock inscriptions, Kartir’s grandee texts, and the Paikuli tower of Narseh (c. 293 CE), whose stone blocks are in the Sulaymaniyah Museum.
Why are two letters named WAW-AYIN-RESH and MEM-QOPH?
Because the Unicode names encode a real merger: in this script one glyph does the work of several Aramaic letters - w/y/r share a shape, and so do m/q. The Inscriptional Parthian block keeps them distinct (22 letters); Pahlavi’s economy of shapes is what later let Book Pahlavi collapse to 13 graphemes for 24 sounds.
How is Inscriptional Pahlavi different from the Book Pahlavi of the texts?
The inscriptional script writes 19 non-joining letters with spaces between words - chiselled, airy, readable. Book Pahlavi, the later manuscript hand, joins its letters into heavy ligatures, shrinks the distinct shapes to 13 for 24 sounds, and leans on context; despite four encoding proposals it remained unsupported by Unicode as of March 2026, so digital editions spell it in romanized transliteration.
Did real people leave their names in this script?
Constantly. A Dura-Europos ostracon lists 27 professions in Middle Persian - baker, shoemaker, plowman, jailer, chief scribe, harem keeper - and on the wall of the city’s synagogue a visitor-scribe signed a doodle: “in the year 15 and the 18th day... the scribe Yazdān-tahm-farrbay, the strong, [came] to this house, and he liked this picture.” Official script, unofficial handwriting.