Languages of antiquity · overview

The family tree of scripts

From Egyptian hieroglyphs to the Latin alphabet, Arabic script and Devanagari

Writing was invented independently only a few times: in Mesopotamia, in Egypt (perhaps prompted by Mesopotamia), in China and in Mesoamerica. Yet most of the world's alphabets – Latin and Cyrillic, the Hebrew and Arabic scripts and, on the prevailing view, the scripts of India – share a single ancestor: signs that Semitic-speaking people borrowed from Egyptian hieroglyphs in the 2nd millennium BCE.

How many times was writing invented?

Writing that can record anything that can be said was, in the view of most scholars, invented independently only a few times. The earliest evidence comes from Mesopotamia, where tablets with pictorial signs appeared at Uruk around 3300–3100 BCE, and from Egypt, where writing appears at the end of the 4th millennium BCE; whether the Egyptians took the idea of writing from Mesopotamia is unresolved. Independently of both, Chinese writing arose, attested from the 13th century BCE, and so did the scripts of Mesoamerica, of which Maya (from the 3rd century BCE) is the best known.

The script of the Harappan civilization in the Indus valley (c. 2600–1900 BCE) remains undeciphered, and it is not even certain that it recorded language. Other scripts arose where people knew of writing from their neighbours: they took over the idea but made their own signs. On the prevailing view, this is also how the first alphabet came about. In the 2nd millennium BCE, Semitic-speaking people in Egypt and Sinai picked a few Egyptian hieroglyphs and used each to write only the first sound of its name: the drawing of an ox head, Semitic ʾalp, stood for ʾ, a house (bayt) for b. Directly or indirectly, most of the world's alphabets descend from this script.

The family of alphabets

The tree shows how the alphabets of Europe, the Middle East, India and Central Asia grew out of Egyptian signs. A solid line means a generally accepted origin. A dashed line means the origin is disputed, or that the script took only a model or an idea from its predecessor rather than the letter shapes. The characters next to each script are a sample, usually its first letters; dates are approximate.

Egyptian hieroglyphslate 4th millennium BCE – 394 CESigns for words, consonants and determinatives. The ox head, house, water and head shown here are, on the prevailing view, the models of the letters ʾalp, bet, mem and reš.
- Proto-Sinaitic and Proto-Canaanitec. 19th–11th century BCEThe earliest alphabet: signs borrowed from hieroglyphs, each writing only the first consonant of its name. Inscriptions from Sinai (Serabit el-Khadim), Egypt (Wadi el-Hol) and Canaan; the date of the earliest is disputed (c. 1850 or as late as 1550 BCE).
  - Ugaritic cuneiform alphabet14th–13th century BCEAbout thirty signs pressed into clay like Mesopotamian cuneiform. The shapes are largely new (a minority of scholars see them as adapted Proto-Canaanite letters), but the letter order nearly matches the Phoenician alphabet, so the principle of the alphabet came from the West Semitic tradition. One tablet also records a second order (h, l, ḥ, m…), the one used by the Ethiopic script.
  - Ancient South Arabianfrom the late 2nd millennium BCE (mature form c. 800 BCE) to the 6th century CEIt split off from the early (Proto-Canaanite) script before the Phoenician alphabet arose, probably in the late 2nd millennium BCE; by what route is unclear. It has 29 consonants. The sample follows the South Semitic order h, l, ḥ, m, q.
    - Ethiopic script (fidäl)with vowels from the 4th century CEAt first consonants only; under King Ezana in the 4th century vowel marks were added and the script became an abugida. Unlike its South Arabian model it runs left to right. Amharic and Tigrinya are still written in it.
  - Phoenician alphabetfrom c. 1050 BCE22 letters for consonants only, written right to left. Phoenician traders carried it across the Mediterranean.
    - Paleo-Hebrew scriptfrom the 10th–9th century BCEA local form of the Phoenician script, used for the earliest Hebrew inscriptions; Unicode does not distinguish it from Phoenician. Among Jews it gave way to the Aramaic script from the Persian period on.
      - Samaritan scriptstill in useA direct descendant of Paleo-Hebrew. The Samaritans still write their Torah in it.
    - Aramaic alphabetfrom the 8th century BCEThe earliest Aramaic inscriptions use the Phoenician script; distinct letter shapes developed by the 8th century. As the script of the Assyrian, Babylonian and Persian administrations it spread from Egypt to India and became the ancestor of most alphabets of Asia.
      - Hebrew square scriptfrom the 3rd century BCEIt grew out of the Aramaic script of the Persian administration, which the Jews adopted in place of Paleo-Hebrew. Hebrew and Yiddish are written in it.
      - Nabataean scriptfrom the 2nd century BCEThe script of the Nabataeans, centred on Petra. Its cursive form gradually turned into the Arabic script.
        Arabic scriptfrom late antiquity (4th–6th century CE)On the prevailing view it grew out of Nabataean, on a minority view out of Syriac. With Islam it spread across much of Asia and Africa; today it is the most widely used alphabet after the Latin one.
      - Palmyrene script1st century BCE – 3rd century CEThe script of the caravan city of Palmyra, known from thousands of inscriptions. It died out after the Romans took the city in 272.
      - Syriac scriptfrom the 1st century CEThe script of Syriac Aramaic, the literary and liturgical language of Middle Eastern Christians. It is still in use.
      - Pahlavi scriptsfrom the 2nd–1st century BCEThe scripts of Parthian and Sasanian Iran. Many words were written in Aramaic but read in Persian (heterograms).
        Avestan scriptfrom the 5th–6th century CECreated under the Sasanians from Book Pahlavi to write down the sacred texts of the Avesta. Unlike its model it records vowels precisely; the sample shows four of them.
      - Sogdian scriptc. 100–1200 CEThe script of Sogdian merchants on the Silk Road in Central Asia.
        Old Uyghur scriptfrom the 9th century CECursive Sogdian adapted for Turkic languages and written in vertical columns.
        Mongolian scriptfrom the 13th century CEAccording to tradition it was adapted from Old Uyghur by Tata-tonga, a Uyghur scribe captured by Genghis Khan. It is still used in Inner Mongolia; the Aramaic branch thus reached as far as China.
      - Kharosthic. 3rd century BCE – 3rd century CEThe script of Gandhara, derived from the Aramaic of the Persian administration; like its model it runs right to left. It attaches vowels to consonants with marks, as Brahmi does.
      - Brahmifrom the 3rd century BCEThe ancestor of most scripts of South and Southeast Asia. Most scholars hold that it was modelled on a Semitic script, probably Aramaic; others argue that it arose in India itself.
        Devanagarifrom the 8th century CEIt developed via the Gupta script and Nagari. Hindi, Marathi, Nepali and Sanskrit are written in it.
        Tamil scriptTamil-Brahmi from the 3rd century BCEFrom the southern branch of Brahmi. The earliest Tamil inscriptions use an adapted Brahmi; the modern letter shapes developed in the Middle Ages.
        Tibetan scriptfrom the 7th century CETibetan tradition attributes it to the minister Thonmi Sambhota under King Songtsen Gampo; according to scholars its model was a late Gupta script of northern India or Nepal.
    - Greek alphabetfrom the 8th century BCEThe Greeks used Phoenician letters for consonants their language lacked (mainly gutturals) to write vowels. The result was the first alphabet to write both consonants and vowels systematically.
      - Etruscan and Old Italic alphabetsfrom c. 700 BCEThe Etruscans took the alphabet from Greeks of Euboea settled in Italy. On the prevailing view the alphabets of other languages of Italy, such as Oscan, Umbrian and Venetic, also came from it.
        Latin alphabetfrom the 7th–6th century BCEOn the prevailing view it came from the Etruscan alphabet. With Rome and later the Church it spread across Europe, and today it is the most widely used script in the world.
        Oghamfrom the 4th century CEAn Irish script of strokes and notches along the edge of a stone. Its shapes come from no other script; most scholars think its set of sounds was modelled on the Latin alphabet, others suggest runes or Greek.
        Runesfrom the 2nd century CEThe model was one of the North Italic alphabets or the Latin alphabet; which one is disputed. The oldest certain inscriptions come from Denmark and northern Germany. The rune row is called the futhark after its first six letters.
      - Gothic alphabet4th–6th century CECreated by Bishop Wulfila for his Bible translation, mostly from the Greek alphabet, with a few letters from Latin and perhaps from the runes.
      - Coptic alphabetfrom the 3rd century CEThe Greek alphabet with six or seven extra letters taken from Demotic (for example ϣ š, ϥ f) for Egyptian sounds that Greek lacked.
      - Glagoliticc. 862–863Created by Constantine (Cyril) for the mission to Moravia. The shapes are distinctive, but on the prevailing view about half of the letters derive from an ornamentally reshaped Greek minuscule; for the rest, Hebrew or Samaritan influence has been suggested.
      - Cyrilliclate 9th – early 10th centuryIt arose in Bulgaria from the Greek alphabet. The letters for Slavic sounds that Greek lacked (such as Ж, Ш, Ц, Ч) were, on the prevailing view, taken from Glagolitic.
      - Armenian alphabetc. 405 CECreated by Mesrop Mashtots. The prevailing view is that its letter order and many letter shapes follow the Greek alphabet; the signs for Armenian sounds that Greek lacked have other sources.
      - Georgian scriptearliest inscriptions 5th century CEIts origin is disputed: the letter order and numerical values match the Greek alphabet, but the shapes are its own. The earliest dated inscriptions are from the 5th century (Bir el Qutt near Bethlehem, Bolnisi). The sample is in the old asomtavruli script.

The tree is simplified: scripts also influenced one another across branches, and scholars disagree about the details of many transitions. Medieval scripts (such as Mongolian) are included only where they show how far the ancient branches reached.

How the alphabet spread and changed

The first alphabets wrote only consonants. That was enough for Semitic languages, where the meaning of a word rests mainly on the consonants of its root: the consonants m-l-k give Phoenician mlk, Hebrew melekh and Arabic malik, ‘king’. Such a script is called an abjad. Vowels later came to be hinted at with the letters for the consonants ʾ, h, w and y, and in Hebrew, Syriac and Arabic eventually also with dots and dashes above and below the letters.

When the Greeks took over the Phoenician script in the 8th century BCE, they used the letters for consonants that Greek lacked (mainly gutturals) to write vowels: ʾalp became alpha, ʿain omicron. This produced an alphabet in the narrow sense, in which vowels stand on an equal footing with consonants. The local Greek alphabets gave rise to the Etruscan alphabet and through it to the Latin one; in late antiquity and the early Middle Ages Greek writing served as the source or model for the Gothic, Coptic and Armenian scripts, Glagolitic and Cyrillic.

Eastwards the alphabet spread mainly in its Aramaic form. Aramaic was the language of administration in the Assyrian, Babylonian and Persian empires, and its script was taken up from Egypt to India. In India a third type of script arose, the abugida: each consonant sign carries the vowel a (𑀓 ka), and other vowels are attached to it as marks (𑀓𑀺 ki). The Ethiopic script works in a similar way; it acquired its vowel marks in the 4th century CE, whether under the influence of Indian scripts is not known.

The order of the letters proved remarkably stable. School tablets from Ugarit of the 14th–13th century BCE have almost the same order as the Phoenician alphabet, and it survived in broad outline in the Hebrew, Greek and Latin scripts: A, B, G (Latin C), D… A second ancient order, beginning h, l, ḥ, m, lives on in the Ethiopic script.

Other roots

Cuneiform, Chinese characters and the Mesoamerican scripts (best known, Maya) belong, with Egyptian hieroglyphs, to the scripts that most scholars consider independent inventions. Anatolian hieroglyphs and the Cretan scripts appeared in a world where writing already existed; an outside stimulus is possible, but their signs are their own. These roots mostly produced scripts for words and syllables, and all of them except Chinese characters have died out.

Cuneiformc. 3300 BCE – 1st century CEIt arose at Uruk from pictorial signs on accounting tablets. The signs wrote words as well as syllables. The script was most likely created for Sumerian (the language of the earliest tablets is disputed), and scribes later adapted it to many other languages of the Near East.
- Akkadian cuneiformfrom the mid-3rd millennium BCEThe Akkadians took over the Sumerian signs and used them to write Akkadian words syllable by syllable, alongside Sumerian signs for whole words. Sample: the syllables a, e, i, u.
  - Hittite cuneiformc. 17th–12th century BCEThe Hittites took cuneiform from Old Babylonian Mesopotamia, probably via northern Syria. Besides Hittite syllables (sample: a, ma, nu, za) they used Sumerograms and Akkadograms.
  - Elamite cuneiformfrom the late 3rd millennium BCEThe Elamites adopted Akkadian cuneiform. In the Persian Empire it was used for administrative records, such as the thousands of tablets from Persepolis.
- Old Persian cuneiform6th–4th century BCEIt looks like cuneiform but is a different system: 36 phonetic signs whose shapes and values do not come from Babylonian writing. Most scholars regard it as a deliberately created new system that took only the idea and the wedge technique from cuneiform; its near-alphabetic structure is often explained by the model of the Aramaic script.
Anatolian (Luwian) hieroglyphsc. 14th–8th century BCEAnatolia's own pictorial script, about five hundred signs for words and syllables, used to write Luwian. It shares only its name with Egyptian hieroglyphs. The sign 𔐀 means ‘I’.
Linear Ac. 1800–1450 BCEThe syllabic script of Minoan Crete. The signs can be roughly sounded out from Linear B, but the language is unknown. Cretan hieroglyphic writing was used on Crete alongside it.
- Linear Bc. 1450–1200 BCEMycenaean scribes adapted Linear A for Greek. Michael Ventris deciphered it in 1952; it died with the Mycenaean palaces. Sample: the syllables a, e, i, o, u.
- Cypro-Minoanc. 16th–11th century BCEThe undeciphered Bronze Age script of Cyprus, thought to derive from Linear A.
  - Cypriot syllabary11th–4th century BCEThe Greeks of Cyprus wrote Greek in it long after the rest of Greece had turned to the alphabet. It is related to Linear B but does not descend from it.
Chinese charactersfrom the 13th century BCEThe earliest evidence is the Shang oracle bones. The only independently invented script still in use. Japan, Korea and Vietnam adopted it, and in Japan it gave rise to the syllabic kana scripts.
Maya scriptfrom the 3rd century BCELogograms and syllabic signs. It belongs to the Mesoamerican writing tradition, which also includes the Zapotec inscriptions of Oaxaca; how the Mesoamerican scripts are related is not settled. Apart from the numerals, it is not yet in Unicode.
Indus scriptc. 2600–1900 BCEShort inscriptions, mostly on seals, about five signs on average. Undeciphered; it is not even certain that they record language. No link to the later Brahmi has been shown. It is not in Unicode.

The roots are not linked to one another because no relationship between them has been demonstrated. Egyptian hieroglyphs head the family tree of alphabets above.

One word, many scripts: king

The word ‘king’ occurs in the texts of almost every ancient culture, which makes it useful for comparing scripts. Where a script had a sign for the whole word, scribes of other languages often wrote it too and read it in their own language. The rows are ordered roughly by the age of the evidence.

language (script)	written	transliteration	note
Sumerian (cuneiform)	𒈗	lugal	literally ‘big man’: lu₂ ‘person’ and gal ‘big’
Egyptian (hieroglyphs)	𓇓𓏏𓈖	nswt	the sw plant is written first but the word is read n-sw-t; in the title nswt-bjtj ‘King of Upper and Lower Egypt’
Akkadian (cuneiform)	𒈗	šarrum	the same Sumerian sign LUGAL, read in Akkadian; the word could also be spelled out in syllables (šar-ru-um)
Hittite (cuneiform)	𒈗	ḫaššuš	again the sign LUGAL as a Sumerogram; the Hittite reading is known from rarer syllabic spellings
Mycenaean Greek (Linear B)	𐀷𐀙𐀏	wa-na-ka	wanax ‘ruler’, Homer’s ἄναξ; qa-si-re-u, the ancestor of later basileus, meant only a local official
Old Chinese (Chinese characters)	王	wáng (modern reading)	the character already appears on the oracle bones; a widely held view is that it depicted an axe blade, a symbol of power
Phoenician (Phoenician alphabet)	𐤌𐤋𐤊	mlk	consonants only; the vowels are supplied from related languages (probably milk)
Hebrew (Hebrew script)	מֶלֶךְ	melekh	the dots and dashes under the letters are vowel points, added only from the early Middle Ages
Aramaic (Aramaic script)	𐡌𐡋𐡊𐡀	malkā	the final -ā (letter ʾ) works as an article: ‘the king’
Old Persian (Old Persian cuneiform)	𐎧𐏁𐎠𐎹𐎰𐎡𐎹	xšāyaθiya	inscriptions may replace it with the logogram 𐏋; hence modern shah
Greek (Greek alphabet)	βασιλεύς	basileús	in Homer still alongside the older ἄναξ
Latin (Latin alphabet)	REX	rēx	related to Sanskrit rājan and to Celtic -rīx in names such as Vercingetorix
Sanskrit (Devanagari)	राजन्	rājan	nominative rājā; in the eastern Prakrit of Ashoka’s edicts lājā
Middle Persian (Pahlavi script)	𐭬𐭫𐭪𐭠	MLKA, read šāh	an Aramaic word written in a Persian text but read in Persian (a heterogram)
Classic Maya (Maya script)	(not in Unicode)	AJAW, a-ja-wa	‘lord, ruler’; kings of the Classic period bore the title kʼuhul ajaw, ‘holy lord’
Gothic (Gothic alphabet)	𐌸𐌹𐌿𐌳𐌰𐌽𐍃	þiudans	from þiuda ‘people, nation’
Geʽez (Ethiopic script)	ንጉሥ	nəguś	hence the Ethiopian title negus; each sign is a syllable
Arabic (Arabic script)	مَلِك	malik	the same Semitic root m-l-k as in Phoenician, Hebrew and Aramaic
Old Norse (Old Danish)	ᚴᚢᚾᚢᚴᛦ	kunukʀ	konungr; so the Jelling runestones (10th century) call kings Gorm and Harald. The Younger Futhark had no separate runes for o and g
Czech (Latin alphabet)	král	–	from the name Karl, after Charlemagne (Proto-Slavic *korľь); Polish król and Russian король have the same origin

Sources and further reading

Peter T. Daniels, William Bright (eds.): The World's Writing Systems. Oxford University Press 1996.
Stephen D. Houston (ed.): The First Writing. Script Invention as History and Process. Cambridge University Press 2004.
Joseph Naveh: Early History of the Alphabet. An Introduction to West Semitic Epigraphy and Palaeography. Magnes Press, Jerusalem 1982.
Andrew Robinson: The Story of Writing. Alphabets, Hieroglyphs and Pictograms. Thames & Hudson, 2nd ed. 2007.
Florian Coulmas: The Writing Systems of the World. Blackwell 1989.