Linear A

Dafato Team | Jun 12, 2022

Table of Content


The "linear A" is a writing, still not deciphered, which was used in ancient Crete. This writing was composed of about ninety signs and ideograms. It is assumed that it transcribes the language of the Minoans. It was Arthur Evans at the beginning of the twentieth century who discovered in Crete the remains of this writing, as well as other ancient writings. He named them "Cretan hieroglyphs", "linear A" and "linear B", according to their appearance and their age. The site of Haghia Triada, in Messara (southern Crete), provided the most clay tablets incised in "linear A".

Two scripts are obviously derived from "Linear A": the "Linear B", used in Crete and Greece, which was deciphered in the 1950s by Michael Ventris and which transcribes a Mycenaean Greek dialect, and the "Cypro-Minoan syllabary", used in Cyprus. The latter is not deciphered, like the "linear A", which would have given the "Cypriot syllabary", which is perfectly legible and notes a Greek language.

The "linear A" dates back to the Minoan period, a period and a civilization of Crete prior to the Greek invasions, around 2000 to 1500 B. It is still used at the same time as the "linear B", but punctually on other sites in particular in southern Crete.

It is usually written from left to right, although rare entries are in the opposite direction. The ISO 15924 code for "linear A" is Lina.

The writing conventionally called "linear A" appears in Crete at the time of the first Minoan palaces, in the Middle Minoan II (between 1900 and 1800, or between 1800 and 1700 BC). It remains used by the Minoan palace administration during the whole period of the second palaces, until the Late Minoan IB (around 1550-1500 BC), and perhaps even until the Late Minoan II (around 1450-1400 BC).

The first documents in "linear A", the documents dating from the time of the first palaces, all come from the palace of Phaistos. They are older than the first documents attested in Cretan hieroglyphs. The period of the second Cretan palaces is the Golden Age of the "Linear A": it is from this period that the immense majority of the documents preserved date, and it is at this time that the extension of the area of dispersion of the documents is maximum. Documents inscribed in "linear A" were found in Crete, but also in the Cyclades, in Kythera and in Laconia.

In contrast to the documents in "linear B", the documents in "linear A" do not only come from palatial centers: they can come from palaces like Knossos or Phaistos, from urban centers like Tylissos, or from sanctuaries like Káto Sými. The diversity of the contexts from which they come echoes the diversity of the types of attested documents, since the "linear A" was used to record both administrative and non-administrative documents, some of which are clearly religious in character. This is another major difference with "shelf B".

The "linear A" ceases to be used when the Mycenaean administration of Knossos supplants the administrations of the Minoan palaces, in MR II. Only an inscription in "linear A" incised on a vase of Knossos, KN Zb 10, could date from this period.

It is possible, but without being able to prove it in the current state of knowledge, that the Summerocretan of the first millennium, written in Greek alphabet, derives from the language noted by the "linear A".

The writing is rediscovered at the beginning of the XXth century of our era, during the excavations of Knossos led by Sir Arthur Evans. The latter succeeded in isolating the "linear A" from the two other Cretan scripts with which it was found, namely the "linear B" and the "Cretan hieroglyphic". Today, the number of inscriptions in "linear A" is about 1500. Attempts to decipher them have remained unsuccessful, despite numerous proposals.

The tablets written in "linear A" are much less neat than their later counterparts in "linear B". They are smaller, and the lines of writing are not separated by horizontal lines. Nor are their contents arranged by entries marking each time the beginning of a new line, but instead it is customary to cut off the words and arrange the results of the operations where there is room, on the other side of the shelf if necessary. This is what makes the analysis of "linear A" shelves so difficult compared to "linear B" ones.

Linear A", in view of the number of known characters, is a syllabic script, like "linear B". The collection of inscriptions in "Linear A" by L. Godart and J. P. Olivier (GORILA), in 5 volumes, provides a standardized presentation of the signs of "Linear A": 178 simple signs are listed (excluding complex signs, excluding fractions), but many signs are observed only once. In fact, the writing seems to use in a current way about 90 signs, of which the majority is found in "linear B".

Indeed, many characters are common between the two scripts, so that it is tempting to "read" Linear A from the known phonetic values of the signs of Linear B. Yet, there are few words in common, but these few common words help to partially validate the hypothesis of the same phonetic value for similar signs in both scripts. For example, the combination of signs read as PA-I-TO in Linear B is found in Linear A, and could mean the city Phaestus.

Following this possible phonetic analogy, the comparison between the two scripts indicates differences in the use of vowels: in "linear B", the five vowels are present: a, e, i, o, u (with the corresponding syllables: da, de, di, do, du, etc.), whereas in "linear A", there is an over-representation of syllables with the vowels a, i and u. This being said, it is possible that, in moving from "linear A" to "linear B", some syllables have changed consonance.

In addition, besides syllabograms, there are many signs interpreted as ideograms or logograms, representing the same objects as in "linear B" (e.g. "barley", "wine", "olives", "oil", etc.). There are also many signs in the shape of a vase

The repertoire of signs

The "linear A" remains undeciphered to this day because of the brevity of the inscriptions found, which seem to be essentially administrative slips. The known inscriptions in "linear A" total some eight thousand signs, whereas it would be necessary to triple this number to be able to carry out serious investigations; by way of comparison, "linear B" was deciphered by Ventris from a corpus of thirty thousand signs.

There have been many proposals for deciphering the alphabet over the last century, but none of them has yet met with the consensus of the scientific community. Most of them are based on the hypothesis that the characters of "linear A" can be read with the phonetic values of similar characters of "linear B" (for the common characters only, knowing that there are many characters in "linear A" which do not have their equivalent in "linear B"). However, this method should be handled with caution, as it is not mandatory that similar signs have the same value in both linears, as illustrated by the comparison of the Latin and Cyrillic alphabets, which share the signs В, С, H, P, Х and {YУ} while assigning them different values.

The hypothesis of belonging to the great family of Indo-European languages is often put forward but is not based on solid arguments.

It has also been proposed that it belongs to the Semitic language family (see below).

In any case, if the phonetic value of "linear A" is the same as "linear B", the transcribed language is not Greek.

The other work carried out, essentially statistical, gave rise to some hypotheses:

Nevertheless, these elements remain for the moment hypotheses.

Names of places

There are names which, when one reads the signs of "linear A" with their "value" in "linear B", correspond more or less exactly to identified places, attested in "linear B":

However, it should be noted that, for the last two places, the forms in "linear A" and "linear B" differ on the vowels. Nevertheless, the probability that they represent well the proposed places is strong.

) at Mount Dicté (found in "linear B" as di-ka-ta-de and di-ka-ta-jo) and I-DA at Mount Ida (however, this last identification is not unanimous since the I could play the role of prefix as in the couple DA-MA-TE and I-DA-MA-TE).

Many other toponyms are found in "Linear B": a-mi-ni-so (Amnisos), a-pa-ta-wa (Aptara), ku-do-ni-ja (Kudonia), e-ko-so (Aksos), ru-ki-to (Lyktos), ka-ta-no (Kantanos), etc., but their identification with groups of signs from "Linear A" is not certain.

Terms of mathematical language

From the contextual elements, the meaning of some words can be proposed. In particular, one of the most frequent words, KU-RO, is located at the end of the tablet, affected by a number that adds up the numbers of the preceding lines. It must therefore mean "total" or a term such as "recapitulation, balance, accumulation, total". Some have proposed a connection with the Semitic term *kwl "all". But other connections have also been made: with the Etruscan churu with a similar meaning, or with the Proto-Indo-European root *kwol "to turn" by metathesis. This term has nothing to do with its equivalent in "Linear B" (to-so), which supports the idea that the language transcribed by "Linear A" is fundamentally different from that transcribed by "Linear B".

The term KI-RO, which often appears in contexts similar to KU-RO, is most likely also an accounting term and

Within the framework of the decimal system (common to the "linear A", the hieroglyphic and the "linear B"), many signs represent numerical fractions, coded by convenience by letters: J, E, F, K, X, A, etc. (for some of them, values were proposed, presenting a good probability of exactitude:





One of the most instructive inscriptions on the values of fractions is HT Zd 156 (found on a wall of the villa of Haghia Triada) where the following sequence appears:

*319 1 *319 1J *319 2E *319 3EF TA-JA K

We can see a geometric series of reason 3

We can assume, if TA-JA is the term for the number 5, as suggested, that TA-JA+K = 81

Votive forms

Although the "linear A" is mainly written on tablets, it is also found engraved on votive objects, with a clearly less utilitarian meaning but probably of a religious nature. A sequence of signs is frequently found, with some variations, in such inscriptions: A-SA-SA-RA, also YA-SA-SA-RA-ME, which is not known to be a title, a god or goddess, or even a prayer.

Recently, an article by Olivier Samson using a multi-sequence alignment of the Votive Form has suggested that Linear A is Minoan Greek and influenced by Semitic deities. The article provides the following translation:

Notably, libations were common in ancient Greek ceremonies and olive oil was a common ingredient. Possibly, the deity SA-RA-ME corresponds to the Semitic goddess, the pole Asherah (Hebrew, singular: As'era ). The deity is sometimes called I-DA-MI ( Greek: δαίμων ).

Other suggestions for word identifications

Other attempts to identify groups of signs have been discussed. The signs MA+RU have been located with a ligature, suggesting that it refers to "wool," evoking its classical Greek synonym ὁ μαλλός (a surprising assonance of this Minoan vocable. MA+RU has been noted with Sumerian bar-LU whose precise meaning is "assembly of the best wool" and which includes the Sumerian logogram bar "fleece" (it is true that this polysemous sign also meant "outside", "entrails", "foreign", "open", etc.). Similarly, another ligature between the signs RU+YA with the possible meaning of "pomegranate" (a fruit dedicated to the great Minoan goddess, like the poppy, and whose seeds played a special role, as in the myth of Persephone) evokes the classical Greek ἡ ῥοιά (hê rhoiá) "pomegranate tree, pomegranate".

The Semitic hypothesis

The Dutch historian and archaeologist Jan Best has proposed that the "Linear A" language belongs to the Semitic language family. In the expression A-SA-SA-RA already mentioned, he found the Semitic goddess Ashera, whose cult was associated according to him with the Minoan labrys. He also wants to have discerned the votive form A-TA-NŪ-TĪ "I have given" as words resembling the northwestern Semitic dialects, i.e. Ugaritic, Phoenician, etc.

The term KU-RO, "total", is close to the synonymous term in Semitic *kwl. Moreover, in one of the tablets of the corpus found at Haghia Triada (HT 31), there is a list of different types of vases with names some of which (if read with the phonetic values of the "linear B") strongly evoke similar names in the Semitic world. But these examples are isolated, and in any case for the names of vases, they may be simple borrowings from another language.

The Indo-Iranian hypothesis

The works published from 1996 by a French researcher, Hubert La Marle, have developed different but convergent methods of deciphering, based both on the comparative epigraphy of the East Mediterranean scripts of the Bronze Age and on the frequencies of the common signs, and have led to the identification of the foundations of a language linked to the Indo-Iranian branch of Indo-European. According to this interpretation, we are not dealing with an agglutinative language but with an inflectional language of the Indo-European type, as had already been assumed by researchers of the Italian school at the end of the 1940s. In principle, the inflections would not have been so different from those of "Linear B", although they are not, in detail, Greek-like endings. H. La Marle presented the results of his work in lectures given at the Faculty of History and Archaeology of the University of Crete (Rethymnon) and in various international meetings.


  1. Linear A
  2. Linéaire A