The Lost Vowels of Proto-Semitic

Imagine trying to read this sentence if it were written “mgn tryng t rd ths sntnc.” You could probably figure it out from context, but what if you’d never heard English spoken before? How would you know it’s “imagine” and not “imogene”, or “sentence” and not “sintince”? This is the exact puzzle that historical linguists face when they stare at the ancient scripts of the Semitic language family.

Languages like Classical Arabic, Biblical Hebrew, Aramaic, and Phoenician were originally written in abjads—writing systems that record consonants but leave most vowels to the reader’s imagination. While this works well for native speakers, it presents a formidable challenge for reconstructing their common ancestor: Proto-Semitic. How do we find the lost vowels of a language that existed thousands of years before its descendants even started writing vowels down? The answer lies in a fascinating story of linguistic detective work, cross-referencing clues from across the ancient Near East.

The Consonantal Skeleton

At the heart of all Semitic languages is the famous triconsonantal root system. A root, typically of three consonants, carries a core concept. For example, the root K-T-B in Arabic and Hebrew carries the meaning of “to write.” By inserting different vowel patterns around and between these consonants, new words are formed:

• Arabic kataba = “he wrote”
• Arabic yiktubu = “he writes”
• Arabic kitāb = “book”
• Hebrew kātab = “he wrote”
• Hebrew mikhtāb = “letter”

Because the consonantal root is so stable and primary, early Semitic scribes saw little need to write the vowels. The context and the grammatical pattern made the meaning clear. The problem for us is that while the K-T-B skeleton is ancient, the vowels in kataba or kātāb are from much later, and they don’t always agree. So which vowels did a speaker of Proto-Semitic use?

Cracking the Code: The Power of Comparison

When a single language isn’t enough, linguists turn to the comparative method. The logic is simple: if two or more related languages share a feature, it’s likely they inherited it from their common ancestor. If they differ in a systematic, predictable way, we can often reconstruct the original sound that caused those different changes.

To reconstruct Proto-Semitic, linguists look for patterns across its major branches:

• East Semitic: Akkadian (the language of ancient Assyria and Babylonia).
• West Semitic:
  • Central: Arabic, Hebrew, Aramaic, Phoenician.
  • South: Ge’ez (Classical Ethiopic) and Modern South Arabian languages.

By comparing words with the same root across these groups, we can start to see patterns emerge. For example, if we have a word that appears with an ‘a’ vowel in several widespread branches, we can be fairly confident the Proto-Semitic word also had an ‘a’ vowel. But the real breakthrough came from a language that broke the mold.

The Akkadian Breakthrough: A Window into the Past

The single most important key to unlocking Proto-Semitic vowels was the discovery and decipherment of Akkadian. Unlike its Semitic cousins, Akkadian wasn’t written with a native Semitic abjad. Instead, the Akkadians adopted cuneiform script from their non-Semitic neighbors, the Sumerians.

Cuneiform is a syllabic script, meaning its symbols represent entire syllables (like ba, bi, bu or ab, ib, ub). To write in cuneiform, Akkadian scribes had no choice but to write down their vowels. Suddenly, linguists had access to a Semitic language from as early as 2500 BCE with its vowels fully preserved in the writing!

Consider the word for “god.” In Hebrew, it’s ʔel (אֵל), and in Arabic, the root is found in ʔilāh (إله). Both point to an original vowel, but it’s hard to be certain. Akkadian, however, gives us the word ilum. The evidence is unambiguous: the word started with an /i/ vowel. From this, we can confidently reconstruct the Proto-Semitic form as *ʔilum.

Another example is “heart.” In Hebrew, it’s lev, and in Arabic, it’s libb. Again, Akkadian provides the tie-breaker: libbum. The proto-form was clearly *libbum.

Reconstructing the Proto-Semitic Vowel Inventory

Thanks primarily to Akkadian, supplemented by evidence from the other branches, linguists have pieced together the basic vowel system of Proto-Semitic. It was beautifully simple, much like the system of Classical Arabic.

There were just three short vowels and three long vowels:

• Short: /a/, /i/, /u/
• Long: /ā/, /ī/, /ū/

With this inventory, we can reconstruct words and understand how they evolved. Let’s take the Proto-Semitic word for “dog”, reconstructed as *kalbum.

• In Akkadian, it remained kalbum.
• In Arabic, it became kalb (losing the final case ending).
• In Hebrew, it became kelev. The change from /a/ to /e/ in this context is a regular, predictable sound shift that occurred later in the Canaanite languages.

Similarly, the Proto-Semitic word for “house”, *baytum, shows another fascinating evolution. In Akkadian, it stayed bītum (where the original *ay* diphthong smoothed into a long *ī*). In Arabic and Hebrew, it became bayt and bayit, respectively, preserving the original diphthong sound.

Adding the Vowels Back In

Centuries after these languages were first written, their speakers also recognized the need for vocalic precision, especially for reciting sacred texts like the Qur’an and the Hebrew Bible without error. This led to the independent development of vocalization systems.

Around the 8th century CE, Jewish scholars known as the Masoretes developed the Tiberian vocalization system for Hebrew—the intricate system of dots and dashes (niqqud) you see in a printed Hebrew Bible today. At roughly the same time, Arab grammarians were finalizing the system of ḥarakāt (the fatḥa, kasra, and ḍamma) to mark the vowels in the Qur’an.

These systems are invaluable, but they represent the pronunciation of their own time, thousands of years after Proto-Semitic was spoken. They are another clue—a very important one—but not a direct photograph of the ancient past. The Hebrew vowel e in kelev (“dog”), for instance, is a later development, and only through comparison with Arabic kalb and Akkadian kalbum do we know the original was an a.

The Resurrection of a Voice

Reconstructing the lost vowels of Proto-Semitic is a triumph of linguistic science. It shows how, by combining different streams of evidence—the grammar of Arabic, the script of ancient Babylonia, the sound changes in Hebrew, and the preservation of old features in Ethiopic—we can reverse-engineer the sound system of a long-dead language. It’s a process of finding missing pieces, solving historical puzzles, and, in a way, giving a voice back to the people who spoke the ancestor of so many of today’s great languages.