Linear B
Michael Ventris and the Decipherment of Mycenaean Greek
- 14 chapters
- 27m
- Ancient History
- Free · no sign-up
Scholars spent decades trying to understand what Linear B meant. Alice Kober developed grids and charts to study the patterns in the tablets. Michael Ventris finally solved the mystery in 1952 when he identified Linear B as an early form of Greek. The tablets recorded everything from inventories of goods to lists of people, showing how the Mycenaean civilization functioned.
The decipherment revealed that Linear B was a syllabic script, meaning each symbol represented a sound. It used both syllabic signs and ideograms, which were picture-based symbols for concepts. While Linear B was cracked, its older relative Linear A remains undeciphered, leaving many questions about Bronze Age Crete. Anyone interested in ancient languages or archaeological discoveries will find this story fascinating.
Listen
-
Read this chapter
Linear B consists of about 200 signs, with roughly half being syllabic and representing sounds, and the rest ideographic, carrying meaning. The process of standardizing how these signs were identified and understood began in Paris in 1956 through international scholarly meetings. By 1961, at the third colloquium held at the Wingspread Conference Center in Racine, Wisconsin, a system largely developed by Emmett L. Bennett Jr. had become widely accepted, known as the Wingspread Convention. That same year, a new organization, the Comité International Permanent des Études Mycéniennes—or CIPEM—was established and later affiliated with UNESCO in 1970 during the fifth colloquium. These gatherings continue today, with the thirteenth one held in Paris in 2010.
Many of the signs used in Linear B are either identical or very similar to those found in Linear A. However, Linear A was used to write a language that remains undeciphered, so it's unclear if the same symbols carried the same sounds in both systems. This uncertainty makes the process of understanding Linear B all the more challenging, as scholars must carefully consider whether similarities between the two scripts reflect actual linguistic connections or mere visual resemblance.
-
Read this chapter
The grid used in the decipherment of Linear B was developed by Michael Ventris and John Chadwick. It shows how the syllabic signs were meant to be read phonetically. The letters "q" in this system stand for labialized velar stops like [ɡʷ, kʷ, kʷʰ], not the uvular sound found in IPA. The letter "j" does represent the voiced palatal approximant [j], which is the sound made by the English letter "y" in words such as "yes" and "yoke." This system helped unlock the meaning of Mycenaean Greek written in Linear B script.
In the study of Linear B, syllabic signs are organized with initial consonants listed in the leftmost column and vowels along the top row beneath the title. Each sign's transcription is given next to it, sometimes uncertain, and is identified by Bennett’s number preceded by an asterisk—following the same convention used by Ventris and Chadwick. These signs appear on tablets and sealings with noticeable variation from one another and from the standardized forms shown. Scholars continue to debate the reasons behind such differences and whether they suggest semantic distinctions.
-
Read this chapter
The signs in Linear B are approximations, each representing around seventy different sound combinations based on rules and conventions. The system uses a grid of monosyllabic signs, mostly of the type V/CV. Some special cases, like 𐁀 (ai) and 𐁁 (au), tested the limits of this model. For example, 𐁀𐀓𐀠𐀴𐀍 spells out Aiguptios, and 𐁁𐀐𐀷 gives Augewās. Still, diphthongs like 𐁉 (rai) are treated as single syllables, such as in 𐀁𐁉𐀺 for elaiwon. Other diphthongs are split into two monosyllables, as seen in 𐀀𐀫𐀄𐀨 for arourans. Vowel lengths and accents aren’t marked.
In Linear B, signs like 𐁌 (Twe), 𐁍 (two), 𐁃 (dwe), 𐁄 (dwo), and 𐁅 (nwa) are thought to begin with labialized consonants rather than two separate sounds, even when they appear alongside two-sign versions. For example, Odatwenta could be written as o-da-twe-ta or o-da-tu-we-ta. Similarly, 𐁈 (rya), 𐁊 (ryo), and 𐁋 (tya) start with palatalized consonants, as seen in forms like -ti-ri-ja for -trja. The sign 𐁇 (pte) is considered an exception by Chadwick, who links its development to pte<*pje, as in kleptei<*klep-jei.
Linear B doesn’t always show the difference between voiced and unvoiced consonants, or between aspirated and unaspirated stops, even though these distinctions were real in Mycenaean Greek. For example, pa-te stands for patēr, while pa-si is phāsi; ko-ru is korus, and ka-ra-we is grāwes. The dental series is different—do-ra represents dōra, showing how voiced dentals are written separately. Sometimes aspiration is marked, but not always: pu-te can mean phutēr, while phu-te-re means phutēres. And initial /h/ only appears before a and rarely—ha-te-ro for hateron, a-ni-ja for hāniai.
In Linear B, the q-series represents labialized velar consonants that no longer existed in classical Greek, having changed into sounds like β, π, φ, δ, τ, or θ over time. These consonants came from Proto-Indo-European roots, assimilation, or borrowed words, especially names. For example, a-pi-qo-ro stands for amphiquoloi, qo-u-ko-ro for guoukoloi, and qa-si-re-u for guasileus, meaning "court official or local chieftain." The j-series shows the semivowel sound like English “y,” used at the start of words or between vowels when the prior syllable ends in i, as in -a-jo for -αῖος and a-te-mi-ti-jo for Artemitios. The w-series also functioned as semivowels, appearing initially or after a syllable ending in u, such as ku-wa-no for kuanos, meaning "blue."
In Linear B, the r-series represents both /r/ and /l/, as seen in ti-ri-po for tripos and tu-ri-so for Tulisos. Some consonants aren’t written but are understood to be present—like word-initial s- and -w before a consonant, such as pe-ma for sperma. The letter pe, usually functioning as its own value in grid class CV, here serves the sound group sper-. This usage followed established rules rather than being innovative. Syllable-final -l, -m, -n, -r, and -s are also omitted, with only word-final velars written plene. For example, a-to-ro-qo represents anthrōquos, where a, primarily of grid class V, functions as an- and could stand for al, am, ar, and so on.
In Linear B, when three or more consonants appear together—like in knosos or khrusos—each consonant is shown by a separate sign that carries its own consonant sound. The vowel in these signs is often described as "empty" or "dead," because it doesn't represent a sound itself. But there were rules for choosing which vowel to use: it usually matched the first vowel of the next syllable, or the one before it at the end of a word. For example, ti-ri-po uses ti- instead of ta- or te- to match -ri-. One unusual case is wanax, where the form wa-na-ka-te appears in the dative, and wa-na-ka-te-ro in the adjectival form, though this may not have applied everywhere. A later example from Agios Vasileios in Laconia shows wa-na-ko-to, written on a sealing nodule from the late 14th or early 13th century BCE.
-
Read this chapter
Ideograms in Linear B function differently from the characters in Han or cuneiform scripts. They carry no phonetic meaning and never serve as word signs within a sentence. Usually found at the end of a line before a number, they indicate the object to which that number refers. Many of their meanings remain uncertain or debated among scholars. Some items like cloth and containers are broken down into various categories, each represented by its own ideogram. Livestock might also be marked according to sex.
The system of numerical references for the ideograms was first created by Ventris and Bennett, who organized them into functional groups based on Bennett's index. These groups start with numbers like 100, 110, 120, and include extra slots for future additions. The numbering system they established is still used today, although the official CIPEM designations now include three- or four-letter codes in small capitals—based on Latin terms that were thought to match the meanings at the time—where those meanings are clear and agreed upon. As of Unicode version 5.0, there are 123 Linear B ideograms encoded.
The ideograms in Linear B are symbolic representations, not literal pictures of the objects they stand for. For instance, a tablet might show a tripod with missing legs, but the symbol used depicts a tripod with three legs. When transcribing these tablets today, scholars often use either English or Latin names—Ventris and Chadwick preferred English, while Bennett favored Latin. Neither system offers a reliable translation of the actual items, and identifying obscure objects often requires careful interpretation.
-
Read this chapter
Tablets bearing Linear B script have turned up on a variety of items, from ordinary tablets and stirrup jars to other objects still being studied. Scholars have identified two main types of these tablets: one kind shaped like palm leaves, where the text runs parallel to the longer edge, and another type that's larger and page-like, split into several lines. The smaller palm-leaf tablets usually contain just one transaction, while the larger page-shaped ones hold summaries of many transactions.
Some 6,058 Linear B inscriptions are now known, including 170 more found on different kinds of containers. The largest group comes from Inscribed Stirrup Jars, which were mostly unearthed at Thebes, Mycenae, Tiryns, and Khania. These markings on pottery and vessels provide insight into how the Mycenaean world managed its affairs, showing the record-keeping practices of that time. Each jar contributes a piece to understanding daily life and governance in the late Bronze Age, as scholars work to uncover what these symbols once meant.
For several decades, scholars have been working to piece together tablet fragments, which has made the tablets and their information more complete while reducing the total number of individual pieces.
The oldest Linear B tablets come from the Room of Chariot Tablets at Knossos, dating to the latter half of the 15th century BC. A clay pebble from Kafkania appears earlier but isn’t genuine. The earliest actual inscription from the mainland is a clay tablet found at Iklaina, dated between 1400 and 1350 BC.
Tablets are organized into series based on the main ideogram found on each one. For instance, the E series includes tablets dealing with grain, marked by the sign gra. Some tablets don’t contain any ideograms at all, and these fall under the X series. Labels and sealings are grouped in the W series, while the Z series covers inscribed objects that aren’t tablets. When researchers can identify further subgroupings within a series—based on shape, vocabulary, format, or handwriting—they add a lowercase letter to distinguish them, like the Er series.
In the year 2000, a discovery was made near Kranzberg in southern Germany, close to a place called Bernstorf. There, archaeologists found an amber seal engraved with signs written in Linear B. This find sparked debate about whether the object was real or not, as its authenticity remains disputed to this day. The seal itself is small but significant, bearing the mysterious script that has long intrigued scholars and researchers alike. Its origins and meaning continue to be discussed among experts, making it a point of interest in the ongoing study of ancient languages and cultures.
-
Read this chapter
Arthur Evans dated the Knossos archive to about 1400 BC, assigning the destruction by fire to the LM II period. This view lasted until Carl Blegen excavated ancient Pylos in 1939 and uncovered Linear B tablets destroyed by fire around 1200 BC, at the end of LHIIIB. When Michael Ventris deciphered Linear B in 1952, doubts emerged about Evans’s dating. Blegen pointed out that the inscribed stirrup jars found at Pylos—similar to those Evans had associated with the 1400 date—were the same type as the oil flasks imported from Crete around 1200. Seeing strong parallels between 1200 BC Pylos and 1400 BC Knossos, Blegen argued the Knossian evidence needed reevaluation, confident in the 1200 BC date he had established for Pylos.
The examination of the Knossos tablets revealed significant problems. Evans had not kept precise records of where the tablets were found. Researchers turned to Duncan Mackenzie’s day books, who had overseen the daily work. But there were differences between Mackenzie’s notes and Evans’s official reports. The two men also disagreed about the exact locations and layers where the tablets had been uncovered. A reinvestigation was later published in a book titled On the Knossos Tablets, which included Leonard Robert Palmer’s essay “The Find-Places of the Knossos Tablets” and John Boardman’s “The Date of the Knossos Tablets.” These represented opposing views held by Blegen and Evans. The disagreement became known as “the Palmer–Boardman dispute,” and it remains unresolved to this day.
-
Read this chapter
Linear B was used by the major cities and palaces of the Mycenaean world to keep records, especially of distributing goods like wool, sheep, and grain, often to groups of religious people. Some tablets also cover military matters, showing how the system recorded both everyday life and defense-related activities. These clay tablets provide a window into the administrative structure of the late Bronze Age, revealing how resources were managed across the region.
When structures containing these tablets were reduced to ash by flames, the heat baked many of them solid, and in that way, preserved them for later discovery.
-
Read this chapter
Plutarch recounts that Agesilaus II, the Spartan king who ruled from 400 to 360 BC, sent a bronze tablet to Egyptian priests. The tablet was covered with "many letters marvellously old," and Agesilaus hoped the priests might understand them, since nothing could be made of them by his own people. This suggests that even in antiquity, the meaning of Linear B remained a mystery, despite its presence in the archaeological record.
-
Read this chapter
Arthur Evans, the British archaeologist who led the Ashmolean Museum, first encountered a Cretan sealstone engraved with writing in 1886, presented to him by Greville Chester. He believed it was Mycenaean, though Heinrich Schliemann had seen similar signs without identifying them clearly, noting only “three or four” combinations in his work on Mycenae. In 1893, Evans bought more sealstones in Athens, learning they came from Crete. The next year, he noticed the script on other objects at the museum. In 1894, he traveled to Crete and found the double axe symbol on a wall at Knossos, which he considered the origin of the script. He also discovered that local women wore similar stones as amulets called γαλόπετρες, thought to boost breast milk production.
Starting in 1894, Arthur Evans began publishing his ideas about Cretan scripts in The Journal of Hellenic Studies. His first article was titled "Primitive Pictographs and a Prae-Phoenician Script from Crete." In it, he separated signs into two categories: "pictographic writing" and "a linear system of writing." He didn’t clearly define these terms, which led to confusion later. By 1898, Evans explained that the linear forms were made of simple geometrical shapes, unlike the more complex pictorial ones, and that they likely came from early, rough drawings—what he called "the infancy of art."
Arthur Evans began negotiating for the land at Knossos, forming the Cretan Exploration Fund with his own money. By 1896, the fund had bought one-fourth of Kephala Hill, where the ruins stood, with first right to purchase the rest. He couldn’t get an excavation permit from the Ottoman government, so he returned to Britain. In January 1897, Crete’s Christian population revolted against the Ottomans. The last Ottoman troops left in December 1898, by which time Evans and his supporters had returned to finalize the purchase. The fund gained new donors by then. In 1899, a new Cretan Republic was established. Once Evans received permission from local authorities, excavation began on 23 March 1900.
On 5 April, excavators working at the site made the first major find of Linear B tablets, uncovering them in a wooden box inside a used terracotta bathtub. The discovery was part of Evans’s report to the British School at Athens that year. Soon after, more caches appeared in various spots, including the Room of the Chariot Tablets, where over 350 pieces came from four boxes. These tablets ranged in size from 4.5 centimeters to nearly 20 centimeters in length and 1.2 centimeters to 7.2 centimeters in width. They had horizontal lines etched into them, and text was written in about 70 characters. Even in this early report, Evans noted that “a certain number of quasi-pictorial characters also occur which seem to have an ideographic or determinative meaning.”
By 2 June, Arthur Evans’s excavation season was finished for the year. He reported that only a small number of tablets were found intact, damaged by rain leaking through the storage room's roof, crumbling, or being discarded by workers who didn’t recognize them. On 6 September, he submitted a report to the Royal Anthropological Institute of Great Britain and Ireland, where he began using ideas that would later define his thinking about Knossos: calling it the “palace of Knossos” and the “palace of Minos.” The report also introduced the term “linear script,” distinguishing it from the older “hieroglyphic or conventionalized pictographic script.” Evans described the linear characters as having a “free, upright, European character” and mostly being syllabic. He also noted that some were unquestionably ideographic or determinative.
After 1900, Arthur Evans was deeply involved in excavations at Knossos, where he discovered tablets and began planning a major work on Cretan scripts titled Scripta Minoa. Just before the first volume of that work was published, he started suggesting publicly that he now thought the linear script actually consisted of two separate scripts, a notion he would present in his upcoming book.
In 1909, Arthur Evans released Scripta Minoa I, noting that the Phaistos Disc discovered in July 1908 had caused him to halt publication so he could include the artifact, since it had not yet been published. On the following page, he mentioned including by Federico Halbherr’s permission unpublished tablets from Hagia Triada, written in a linear script labeled “Class A.” It is unclear whether Halbherr influenced Evans’s division of the linear script into Class A and Class B. The Knossos tablets were classified as Class B, suggesting that Evans recognized Class A only in materials from other sites, recent enough that permission was needed to include them.
Arthur Evans classified the Linear scripts based on how the symbols were formed and arranged, which he described as "type" or "form of script." He noted differences between the two, calling them Class A and Class B. For example, he says "the clay documents belonging to Class A show a certain approximation in their forms," indicating that the characters in one class had more consistent shapes than those in the other. This distinction helped lay the groundwork for later efforts to understand these ancient writings.
-
Read this chapter
In the years following the discovery of the Cretan script, scholars began attempting to understand what it might mean. One such effort came from Florence Stawell, an Australian classicist, who in 1911 shared her interpretation of the Phaistos Disc in The Burlington Magazine. She later expanded on this work in a 1931 book titled A Clue to the Cretan Scripts. Stawell argued that all three forms of the Cretan script represented early Homeric Greek and offered her own translations. Also in 1931, F. G. Gordon published Through Basque to Minoan, proposing a link between the Basque language and Linear B, though his theory did not gain lasting acceptance.
In 1949, Bedřich Hrozný published a work titled Les Inscriptions Crétoises, Essai de déchiffrement, offering a proposed decipherment of the Cretan scripts. Hrozný was internationally known for his earlier translation of Hittite cuneiform. His attempt to translate the Minoan texts into academic French, however, turned out to be quite subjective and ultimately incorrect.
From the 1930s to 1950s, researchers across the globe shared their findings about Linear B through letters and published papers. Among them were Johannes Sundwall, K. D. Ktistopoulos, Ernst Sittig, and V. I. Georgiev. Though none of these scholars succeeded in unlocking the script’s meaning, their contributions helped build the ongoing scholarly conversation around this mysterious ancient writing system.
-
Read this chapter
About the same time, Alice Kober turned her attention to Linear B and began organizing the symbols into groups of three, noticing patterns that linked similar signs. She observed that many Linear B words shared common roots and suffixes, which led her to conclude the script represented an inflected language, where nouns changed their endings based on case. Yet some characters in the middle of words didn't seem to fit with either root or suffix. Drawing on what she knew about other languages, Kober guessed these were bridging syllables—where the start belonged to the root and the end to the suffix. This made sense because Linear B had too many signs to be purely alphabetic and too few to be logographic, meaning each symbol likely stood for a syllable. Her methodical work revealed three grammatical cases and helped identify pairs of signs that shared vowels or consonants.
Alice Kober noticed that in the Linear B script, the two-symbol word for 'total' used at the end of livestock and personnel lists changed its symbol depending on gender. She observed that this kind of gender shift, brought about by altering just one letter—usually a vowel—was especially common in Indo-European languages. Though she avoided guessing what language Linear A or Linear B might represent, Kober did think it very likely that the two scripts stood for different languages. Her focus remained on carefully cataloging and analyzing the actual symbols themselves.
-
Read this chapter
Emmett L. Bennett Jr. created the numbering system still used today for Linear B symbols. Working with Alice Kober, by 1950 Bennett had figured out the metrical system from studying Pylos tablets. He determined those tablets used the same script as Linear B found at Knossos. As he prepared a publication on the Pylos tablets, Bennett classified and numbered the signs. Like Kober, he believed Linear A and B were different languages. His book The Pylos Tablets became essential reading for Michael Ventris, who later called it "a wonderful piece of work."
-
Read this chapter
In 1935, Arthur Evans spoke at an exhibition held by the British School at Athens in London, where a young Michael Ventris was in attendance. By 1940, Ventris had already published an article titled “Introducing the Minoan Language” in the American Journal of Archaeology. After serving as a navigator with RAF Bomber Command during the war and spending time in occupied Germany afterward, he returned to civilian life, completed his training as an architect, and kept up his study of Linear B, writing to scholars who sometimes responded.
Michael Ventris, working with John Chadwick, a university lecturer in Ancient Greek philology, made the breakthrough in deciphering Linear B between 1951 and 1953. Initially using his own numbering system but later adopting Bennett's method, he began by applying Kober's classification tables and his own theories. Noticing that some symbol combinations appeared only on tablets found in Crete, he guessed they might be place names—correctly. As he decoded more text, Ventris realized the language was Greek, contradicting contemporary views and even his earlier agreement with Evans's belief that Linear B wasn't Greek. On 1 July 1952, Ventris announced his discovery on BBC Radio, calling it "a difficult and archaic Greek." Chadwick, who collaborated closely with Ventris, later noted that no evidence had ever shown Linear B was anything other than Greek, and that unusual words were to be expected in such an ancient form of the language.
After Michael Ventris died in a car crash, the first version of their book Documents in Mycenaean Greek came out in 1956.
Michael Ventris’s work decoding Linear B didn’t win immediate acceptance, and some scholars were unsure of his conclusions. Professor A. J. Beattie of Edinburgh raised questions about it in the later 1950s. Saul Levin from the State University of New York thought that although Linear B had Greek features, it also included an earlier substrate, a theory he developed in his 1964 book The Linear B controversy reexamined. But by the mid-1950s, many researchers began to support Ventris’s findings. Scholars like Carl Blegen and Sterling Dow backed his approach, and together with Ventris’s 1954 article, this helped lead to broader scholarly approval of his breakthrough.
-
Read this chapter
In April 2003, Linear B became part of the Unicode Standard with the release of version 4.0. The encoding for Linear B uses specific blocks within Unicode: the Linear B Syllabary spans U+10000 through U+1007F, while the Linear B Ideograms are found in U+10080 to U+100FF. There is also a separate block designated for Aegean Numbers, which connects to this system and ranges from U+10100 to U+1013F. Various fonts now support the display of Linear B text.
Read
Free to download, keep and share. For general information only — not professional medical, legal or financial advice. Please consult a qualified professional.