1. How Many Scripts Are There, and Why Sort Them

The most practical count of the world’s scripts comes from Unicode, the international standard computers use. On 9 September 2025 the Unicode Consortium released version 17.0. It added 4,803 new characters, for a total of 159,801, and brought the number of scripts to 172.

Four scripts were newly added. Beria Erfe is a modern script used by the Zaghawa community of central Africa. Tolong Siki is the script of the Kurukh community of northeast India. Tai Yo is the traditional script of the Tai Yo community of northern Vietnam. Sidetic was used in ancient Anatolia. Scripts are not museum pieces. They are living tools that are still being created and standardized.

A list of 172 cannot be memorized, but sorted by how they write sound, the scripts fall into a few families. Alphabets write consonants and vowels as separate letters. Abjads write consonants and largely omit vowels. Abugidas build a default vowel into each consonant letter and change it with marks. Syllabaries write one syllable per character. Logographic scripts use units of meaning as characters. This sorting says nothing about better or worse. It shows the different solutions people chose to fit the sound structure of their languages.

2. The Birth of the Alphabet: From Hundreds of Signs to Just Over Twenty

According to the Penn Museum, the alphabet was invented by the Canaanites, the predecessors of the Phoenicians. They took their inspiration from Egyptian hieroglyphs, possibly as early as around 1800 BCE. Clay tags recently found at Umm el-Marra in Syria may push the date back as far as 2400–2300 BCE.

The alphabet’s decisive advantage was number. Cuneiform and hieroglyphs used hundreds of signs. By contrast, “students of alphabetic writing only need to learn twenty to thirty letters.”

The same source notes that the earliest Greek alphabetic inscriptions date to the 700s BCE. This was when Phoenician ships sailed the Mediterranean. The museum also notes that the Latin alphabet developed from the Greek alphabet. Today the Latin alphabet is used for English and many other languages.

It is widely known that Phoenician writing recorded only consonants. It is also widely known that the Greeks reassigned some letters to vowels. However, the museum material checked for this article does not cover that process in detail, so we note only the outcome. Greek and Latin became scripts that write both consonants and vowels as separate letters. That approach has become the most widely used type of writing in the world today.

Woodblock-printed page of the Hunminjeongeum Haeryebon explaining the articulatory design principles of the Korean Hangul script
A page from the 1446 Hunminjeongeum Haeryebon (UNESCO Memory of the World), explaining how Hangul letters model the human speech organs and combine into syllabic blocks · Kbarends at English Wikipedia · Public domain · Wikimedia Commons

This shows that the alphabet's date of origin may be pushed further back by new finds. Its strength was the small number of letters, but that can be read as a practical advantage, not a sign that one script is better than another.

3. Abjad: Why Arabic Leaves Out Its Vowels

Arabic is the best-known example of a consonant-centered script, called an abjad. Richard Ishida, who works on internationalization at the W3C, maintains notes on Arabic orthography. They describe the script as an abjad in which “short vowels are not normally written.”

Arabic has 28 basic consonant letters. It is written horizontally from right to left. However, numbers and embedded Latin text are read left to right, so the direction can change within a single line. The script is always cursive. Letter shapes change depending on what comes before and after them.

This does not mean Arabic has no vowels. Long vowels are indicated by three letters, alif, waw, and ya, which double as vowel carriers. Short vowels can be written with small marks above or below the letters, called harakat.

Those marks, however, appear only in “Qur’anic texts, dictionaries, educational materials, and where the pronunciation needs to be made clear.” Most everyday print omits them. The reader knows the language and recognizes words from the consonant skeleton.

The same script also writes Persian, Urdu, and many African languages, adapted with extra letters for each language’s sounds. This is a reminder that scripts and languages do not map one to one.

Set beside the alphabet, the difference between the two types turns out to lie less in whether vowels exist than in whether they are always written. That vowel marks appear in dictionaries and teaching materials shows that leaving them out is a choice that assumes a reader who knows the language.

4. Abugida: The Vowel Lives Inside the Consonant

Devanagari and the other scripts of South Asia belong to a type that a Unicode technical note calls “abugidas or alpha-syllabaries.” In this type, the note explains, “consonant characters represent a consonant+vowel syllable.” Each consonant carries an inherent vowel, a built-in vowel that must be overridden when a different vowel is needed. The pronunciation of that inherent vowel varies by script, with examples such as /ə/, /ʌ/, and /ɔ/.

To write a vowel other than the inherent one, a vowel sign is added. In Sanskrit this sign is called a matra. It goes to the left, right, above, or below the consonant, sometimes on more than one side at once. To pronounce a bare consonant, a small mark is used. Unicode calls it a virama, and it “kills” the inherent vowel.

When consonants cluster, their shapes change or merge. In 60 percent of Devanagari conjuncts, a half-form is used. In this half-form, the vertical stroke historically linked to the inherent vowel sound disappears. All these scripts, the note says, descend from a common ancestor, Brahmi.

Africa has its own abugida, Ethiopic. The Unicode core specification describes it as a script with the tradition of “presenting combinations of consonants and vowels as a two-dimensional matrix.” Its encoding follows that structure: 43 consonants crossed with 8 vowels, making 344 conceptual syllables.

Ge’ez, the language that gave the script its name, now survives only as a liturgical language. But the script itself has been adopted for Amharic, Tigre, Oromo, and other languages of East Africa.

The same document also records scripts created in Africa in modern times. N’Ko is a right-to-left script devised in 1949 to write the Manden-family languages of West Africa. The Vai syllabary of Liberia and Sierra Leone was created in the 1830s, and a standard syllabary chart was published in 1962. The Bamum syllabary was created between 1896 and 1910 in western Cameroon.

Classical illuminated Arabic manuscript showcasing an abjad writing system where consonant roots carry the primary script structure
Classical Arabic calligraphy—an example of an abjad script system structured around consonant roots, with optional diacritics for short vowels · Calligrapher: Abu Muhammad Khan al-Ma'rashi · Public domain · Wikimedia Commons

Compared with the abjad, the abugida can be read as a solution that does not drop the vowel but sets one as the default. Placing Devanagari beside the Ethiopic script shows that even within one type, the ways of attaching and arranging vowels differ.

5. Syllabary and Logographs Together: Japanese

Japanese uses three kinds of characters within a single sentence. Kanji write units of meaning. Two syllabaries, hiragana and katakana, write one syllable per character.

How many kanji does one need? Japan’s Agency for Cultural Affairs lists 2,136 characters in the Jōyō Kanji Table, which was revised by Cabinet notification in 2010 (Heisei 22). The agency describes the table as a “guideline (目安) for kanji use in general social life.”

In other words, the jōyō list is not every kanji required by law. It is a benchmark set at the level considered sufficient for ordinary life, such as official documents and newspapers.

Each kana set is commonly described as having 46 basic characters that write the syllables of Japanese. Hiragana is used mainly for grammatical elements such as particles and endings. Katakana is used for loanwords and similar purposes. Here one language divides labor between a logographic script and two syllabaries. This shows that script types are not assigned one per language; they can be combined as needed.

Describing the list as a "guideline" shows that the 2,136 characters cannot be taken as the whole of the kanji used in Japanese. Seen together with kana, the load of writing Japanese can be read as spread across the whole combination rather than resting on a single script type.

Fragment of an old scroll with waka poems in flowing vertical hiragana
Koyagire fragment: opening of Volume 9 of the Kokin Wakashū · Unknown author Unknown author · Public domain · Wikimedia Commons

6. Hangeul: Letters Modeled on the Organs of Speech

Hangeul, the script of Korean, writes consonants and vowels as separate letters, like an alphabet. But it holds a special place in the history of writing, because the record of how and why it was designed survives.

Korea’s National Institute of Korean History records that King Sejong created 28 letters in 1443, the 25th year of his reign. In the ninth month of 1446 he published the Hunminjeongeum, known as the Haerye edition. The National Hangeul Museum’s permanent exhibition is titled “Hunminjeongeum: A Thousand-Year Writing System Plan.” It explains that those 28 letters comprised 17 consonants and 11 vowels. The basic consonants were modeled on the shapes of the speech organs. The vowels were derived from the principles of heaven, earth, and humanity.

UNESCO inscribed the Haerye edition on the Memory of the World Register in 1997. UNESCO describes it as the book published in the ninth lunar month of 1446 that promulgated the script Sejong had completed in 1443. The book also contains commentaries and examples by scholars of the Hall of Worthies. The original is held by the Kansong Art Museum.

The Hangeul used today does not have 28 letters. According to the Encyclopedia of Korean Culture’s entry on jamo, current Korean orthographic norms set the basic letters at 24: 14 consonants and 10 vowels.

Four letters fell out of use as the sounds they wrote disappeared or merged. The arae-a (ㆍ) wrote a fifteenth-century back low vowel. The yeorin-hieut (ㆆ) wrote a glottal stop. The yet-ieung (ㆁ) wrote a velar nasal. The banchieum (ㅿ) wrote an alveolar fricative. A script keeps changing after it is created, following the sounds of its language. Hangeul’s 24 letters are the result of that change.

Traditional Sanskrit manuscript written in the Devanagari abugida script with horizontal top bars (shirorekha)
A Sanskrit manuscript in Devanagari, an abugida writing system in which each consonant letter carries an inherent vowel modified by attached vowel signs · Ms Sarah Welch · CC BY-SA 4.0 · Wikimedia Commons

This makes clear that 28 and 24 are not conflicting records but figures from different points in time. Set beside Japan's Jōyō Kanji list, it also reveals a shared feature: today's letter count is fixed by norms that government bodies publish.

▶ Official YouTube Video
국가유산채널(K-Heritage Channel)
Hunminjeongeum Part 1: The Hangul Commentary, Hunminjeongeum Haerye Watch the video on YouTube →

7. Scripts and Languages Are Not the Same Thing

One fact runs through every example above: scripts and languages are different things. The Arabic script writes not only Arabic but also Persian and Urdu. The Ethiopic script went on writing Amharic and Oromo after Ge’ez retreated to the liturgy. Conversely, Japanese is one language written with several scripts. So the phrase “the script of such-and-such country” should always be used with care. A script is a vessel for the sounds of a language. Vessels can be borrowed, modified, and newly made.

And they are still being made. Consider N’Ko in 1949, Vai in the 1830s, and Bamum in the early twentieth century. Add Beria Erfe and Tolong Siki, newly encoded in Unicode in 2025. All are scripts that communities created in the modern era to write their own languages. The assumption that older scripts are more legitimate does not fit this living history. Hangeul carries a documented date and design rationale from 1443. In the same way, new scripts are born with their own reasons and designs and take their place in the standard.

Through this distinction, it becomes clear that Unicode's 172 is a count of scripts, not of languages. It also follows that the number of scripts and the number of languages cannot be counted as the same thing.

Dark stone slab densely inscribed with text, displayed in a glass case
The Rosetta Stone on display at the British Museum, 2026 · APK · CC BY 4.0 · Wikimedia Commons

8. Three Things to Remember When Meeting an Unfamiliar Script

First, do not judge a script’s difficulty or merit by its letter count. Arabic has 28 letters but requires learning vowel omission and joining rules; Devanagari’s core is the combination of marks; kanji has the 2,136-character jōyō benchmark. Each script has simply placed its complexity in a different spot to fit its language.

Second, distinguish the name of the script from the name of the language. Saying “Urdu written in the Arabic script” or “Amharic written in the Ethiopic script” reduces misunderstanding. It is for the same reason that the encoding and fonts of digital text are handled by script rather than by language.

Third, check the standards. The Unicode Consortium’s public materials confirm whether a script is encoded in Unicode and which script family it belongs to. National norms are published by government bodies, such as Korea’s orthographic rules or Japan’s Jōyō Kanji Table. There are roughly four ways to write sound. But the histories of the people who chose them are inscribed differently in each of the 172 scripts.