inaugural date: 30 November 2000; latest update: 3 July 2023
Comments, corrections, questions: John Younger (email@example.com)
1. List of Linked Files
The following fonts are now available (7 Sep 08) for Macintosh OS X (courtesy Jean-Pierre Olivier):
3. Corpora and Phonetic Transcriptions
The transcribed texts are based on the texts presented in GORILA (below). With the permission of Jean-Pierre Olivier and Louis Godart, I have prepared a transnumeration of the texts (converting GORILA's drawn signs to their corresponding sign-numbers) and from this transcription I then prepared a phonetic normalization of the texts, using the conventional phonetic values assigned to Linear A signs that correspond to the known Linear B signs. This process was completed on 22 March 1994. Olivier graciously checked this document against GORILA vols. I-V and a ms. of VI. It was then put in tabular form in January-February 1997. Since then, there have been continual updates.
The phonetic transcriptions use Linear B values for Linear A signs assumed to be the same. Also see below, "Phonetic values of the signs."
4. Conventions (definition of "inscription"; bibliographical &
I am NOT interested in producing a decipherment of Linear A -- please see "Decipherments" (section 8, below).
The first sign looks Hieroglyphic, like bull head Hieroglyphi sign *012 ; the other two signs are definitely Linear A: TA-JE. If the inscription is indeed a mix of scripts, it may well be very early.
The other earliest documents date to MM IIA (KN 40 from Knossos, South House, carrying a badly legible fraction) or MM II (ARKH Zf 9; PH 6-19, 22, 24-28, 30 [Haghia Photini], Wb 33-36, Wc 37-41, 43, 44, 46, 52, 55, Wg 45, and Wy 42; and SAM Wa 1).
Hieroglyphic was therefore probably invented first, in MM IA and appears first on seals from Archanes and Ayia Triada; Linear A follows immediately in MM IB, or soon after, in MM II, and appears first on documents primarily from Phaistos. From then until MM III, Hieroglyphic and Linear A were being written contemporaneously, with Hieroglyphic documents at Malia Quartier Mu (MM II) and Malia Palace (MM III, and Knossos Palace?) and with Linear A documents at Phaistos (MM II), Malia Palace (MM III), and Knossos Palace (MM IIIB). From this evidence, it is possible that Hieroglyphic originated in central Crete first (or possibly at Malia), in MM IA and Linear A originated at Phaistos slightly later in MM II.
Although the two scripts share several signs, which may have similar phonetic values, it is not clear why two such different scripts should have developed more or less contemporaneously unless they represent two different administrative practices and/or two different languages or dialects (Schoep 2002, 22-23).
Other linear scripts may have similarly developed from Linear A farther east: see the inscription from Lachish (Finkelberg 1996).
|Linear A||Linear B||Greek||meaning|
|SI+AU+RE||si-a2-ro||σίαλος||to be fattened|
|Linear B||Linear A|
|*76 ra2 /rja/||PA-SA-RI-JA (HT 24a.4), KU-PA3-RI-JA (HT 24a.1)|
|a-pu-wa (TH Fq 229.3)||A-SU-PU-WA (ARKH 2.5-6)|
|*86 dwa??||DA-WA-[•]-DU-WA-TO (KN Za 10a-b)|
|*19 ru2?? /rju/||*309a+RI-JU (logogram on TY 2.3, .6)|
|*92 sja?? (PY Ma 397[+]1048.1: a-•-ta2 = a-*92-ta2)||A-SI-JA-KA (HT 28.1, b.1-2) and common ending -SI-JA-SE|
|NAME||a: contribution||b: U-MI-NA-SI||assessment|
|SA-RA2||OLE+DI 1||OLE+DI 5||6|
|NI 2||NI 2||4|
|VINa 3||VINa 4||7|
|VIR+KA VINa 6||6|
|VINa 3 E||3 E|
|OLE+?||JA-QIf||3 J L2||3 J L2|
The ratios seem to be as follows:
|name||a: KI-RI-TA2 (owed)||b: SA (paid?)||assessment|
|SA-RA2||a: GRA 10||10|
|a: VINa 1||b: SA (paid?) VINa 9||10|
|11 oil, figs, cattle|
Again, the ratios appear to be similar in proportion to those in HT 28:
This is not to say that the acrophonic principle is never appropriate to Linear A. Valério 2007 demonstrates that the word for master/lord is DU-PU2-RE and that the first sign DU is based in form on the Egyptian sr , "official/dignitary/courtier."
Thus, I am not persuaded that Linear A is "such and such" a language because some of its vocabulary looks borrowed. Some scholars have identified the Linear A word KU-NI-SU as Semitic for "emmer wheat" (Gordon 1975 and 1981, Best 1988). It may be but it does not function as a commodity in the three documents it appears in: HT 10a.1, as a heading and therefore probably a NAME; and HT 86a.1-2 and HT 95a.3-4 & b.3-4 listing the same 6 NAMES contributing grain, among which is KU-NI-SU — it is probably an important placename. Compare Nakassis and Pluta 2003: 335: "A number of scholars have attempted to decipher Linear A, identifying it with known languages such as Semitic, Luwian, and even Greek. These studies begin by attempting to etymologize a small number of individual words, largely ignoring overall context."
The contextual method My own method has been strictly internal and contextual, to examine the texts as accounting documents and to use the numbers to identify transaction terms and patterns in vocabulary. The results of these investigations suggest that Minoan is a non-Indo-European, non-Semitic language with a largely standard set of phonemes, a morphology (perhaps with agglutinating tendencies) that incorporates both prefixes and suffixes, and (perhaps) VSO word order.
Yves Duhoux (1978) has paid special attention to vocabulary variations, especially in prefixes and suffixes. Linear A is heavily prefixed, unlike Indo-European languages which are heavily suffixed (with declension and conjugation endings). Linear A is thus not likely to be an I-E language.
Brent Davis (2011/2014) has established some principle characteristics of Linear A: its word order conventionally follows verb-subject-object order.
"The languages which have been used for comparison are of two families: Indo-European, especially an Anatolian language such as Luwian (Palmer, Meriggi [and Ed Brown of UNC-CH]); Semitic (Gordon, Best, and others)... First no inflexional forms such as characterize Indo-European or Semitic languages can be clearly demonstrated, hence the identifications depend largely on vocabulary, which is notoriously easily borrowed. Secondly, the Semitic comparisons are mainly with triconsonantal roots -- yet if the vowels are ignored we are leaving out half the information presented by the script, and thus much decreasing the chances of success. Thirdly, if the language of Linear A does not belong to a well-known family, then the chances of identifiying it are virtually nil. This is not to say that Linear A remains undecipherable; as more documents are found and published, we shall understand more of it. But I doubt very much if speculation at this stage can help; I feel strongly that it is likely to belong to an unfamiliar type." (Chadwick 1975: 147)
Phonetic values of the signs (Godart 1984, amplifying Olivier's previous list)
Many of the Linear A signs were adopted by Linear B; their phonetic values are assumed to be the same.
*22=mPI2 (see Duhoux 1984, Janda 1986, Melena 1987; Tosa 2010)
*29=mPU2 (see Duhoux 1984, Janda 1986, Melena 1987; Tosa 2010)
*56=PA3 or mPA3 (see SMID 1981, p. 61; Duhoux 1984, Janda 1986, Melena 1987; Tosa 2010)
*65=JU (see SMID 1981, p. 61)
*66=TA2=TNA (Pope-Raison 1978: 28).
*304 = KA2
*306 = A2, always in initial position (shape resembles AB 43, known from MY Zf 2)
*315 = RO?
*318 = DI2
|*304 = KA|
|JA-*304[ (PH 14a)||cf. A-SI-JA-KA (HT 28a.1, 28b.1-2)|
|*304+PA (lots)||KA-PA (HT 6a.1; HT 8b.4; HT 94a.1; HT 102.1; HT 140.5)|
|*304+PA+*316+D3 (HT Wa <1021bis>)|
|*304+PA-KU-PA (HT We 1020a)|
|*306 = A|
|]*306-JA-PI (ARKH 3b.1)||WA-JA-PI-[ ] (HT 9b.1)|
|]*306-KI-TA2 (HT 122b.2)||A-*301-KI-TA-A (TY Zb 4)|
|]*306-QE-DU[ (KH 21.3)|
|]-*306-TI-KA-A-RE[ (HT 4.1)||A-TI-KA (ZA Wc.a1-2)|
|*306-TU-JA (HT 115b.3)||cf. JA-TO-JA[ (ZA 4a.2-3)|
*314 = PU3.
10a. Transaction Signs
In Linear B, transaction signs can mean different things in different contexts (e.g., DA in PY En 609 or in Aa, Ab sets).
In Linear A,Transaction Signs "provided information about the kind (or purpose) of transaction and that the format of a tablet indicated the use to which a commodity was put by the administration (its administrative status), its provenance or its destination. Thus, the lists of mixed commodities are clearly drawn up to a different administrative purpose than the single commodity tablets: the former record a single origin or destination for many commodities whereas the latter record several origins or destinations for single commodities" (Schoep 2002, 141).
The low frequency of most transaction signs implies they are context-bound, which would be appropriate for contributions; those occurring with KI-RO (e.g., HT 123) probably are also dealing with "incoming commodities or assessments of incoming commodities" (Schoep 2002, 142) that have not been fulfilled or delivered..
KA, apparently a container (see HT 93) (Schoep 2002, 137, identified it as something to do with people or a group of people; cf. HT 11b.2). The following discussion explores the possibility that VIR+KA are porters (see the individual notes to HT 28 and HT 88, and the note to HT 93 that identifies KA as containers).
10b. Transaction Words
10c. Place Names
Otherwise, SA-RA2 appears as the heading to a list of commodities, presumably contributions "to" SA-RA2.
Some 12 tablets (HT 18, 28a+b, 30, 90, 94, 99, 100, 101, 114, 121, 125, 130) associate SA-RA2 with a set of 4 commodities (*303, FIC, VINa, GRA, OLE), usually in that order when they occur together. But SA-RA2 is occasionally listed against other commodities: BOSm (HT 114a.3-4), VIR+KA HT 28a.4). HT 30, 32, 33, 34 seem to form a separate set of SA-RA2 documents, listing QA+[?]+PU, QA2+[?]+RE, MI+JA+RU, MI+JA+KA, E+KA, *305, *341+PI, PA3+QE.
That SA-RA2 may be Ayia Triada itself is implied by HT 97.b where a carelessly written SA-RA2 occupies the entire side (see commentary to HT 97b). HT Wc 3017 may refer to HT 94 (see the commentary there); if so, its retention at HT may reinforce the identification of HT as SA-RA2..
10d. Other Words
The Libation Formula appears in part on various inscribed objects, many dedicated at sanctuaries. The words, however, often follow a fairly strict sequence, something like the following:
|sequence 1||sequence 2||sequence 3||sequence 4||sequence 5||sequence 6|
|TL Za 1|
|A-TA- I-*301 -WA-JA||O-SU-QA-RE||JA-SA-SA-RA-ME||U-NA-KA-NA-SI||I-PI-NA-MA||SI-RU-TE|
|IO Za 6|
|TA-NA- I-*301 -U-TI-NU||I-NA-TA-I-ZU-DI-SI-KA||JA-SA-SA-RA-ME|
Apparently sequences 1,2,3 can stand on their own, while sequences 4,5,6 form a second clause, which, however, never stands on its own.
A varient of Sequence 1 also appears on ZA Zb 3, a pithos found, not in a sanctuary but in a rural farm house (Ano Zakros):
This suggests that -I-*301 is the root with prefixes and suffixes. The fact this word thus appears on a pithos in a house also suggests that this important word is a common one. Davis suggests a verb, like "give."
In which case, perhaps Minoan was a verb-initial language.
Since sequence 2 is always different, that should be the name of the person "giving."
Thus, Minoan is a verb-initial / subject-second. That would then put the object of the verb third: VSO.
If so, JA-SA-SA-RA-ME, sequence 3, should be the object, the inscribed object itself, the "gift/dedication."
Sequences 4,5,6, comprise a second clause that never varies but never stands alone; they are thus generic, like a generic dependent clause that restates the main clause; Davis suggests something like "requesting favor divine" (in many VSO languages, adjectives follow their noun).
Thus, the Libation Formula would translate, in most cases, something like the following:
|sequence 1||sequence 2||sequence 3||sequence 4||sequence 5||sequence6|
|TL Za 1|
|A-TA- I-*301 -WA-JA||O-SU-QA-RE||JA-SA-SA-RA-ME||U-NA-KA-NA-SI||I-PI-NA-MA||SI-RU-TE|
|gives||O-SU-QA-RE (name of dedicant)||this dedication||requesting||a favor||divine|
The pithos, ZA Zb 3, cited above, has the following inscription:
|WINE 32 (units)||DI-DI-KA-SE||A-SA-MU-NE||A-SE||A-TA-I-*301-DE-KA||A-RE-PI-RE-NA||TI-TI-KU|
Here, it is the OBJECT (the wine pithos) that is placed in initial position as emphasis. This implies an inversion of the typical order: OVS.
We know the verb (A-TA-I-*301-DE-KA), following which should now be the subject A-RE-PI-RE-NA TI-TI-KU.
This implies that the phrase DI-DI-KA-SE A-SA-MU-NE A-SE comprises another sort of object before the verb, perhaps an indirect object:
|wine-pithos||(to) DI-DI-KA-SE A-SA-MU-NE A-SE||(verb) A-TA-I-*301-DE-KA (gives/gave)||(the subject, the giver) A-RE-PI-RE-NA TI-TI-KU|
Other Nouns and other parts-of-speech can occur in the concise headings to accounting documents, but their conventional layout and narrow focus do not encourage elaboration or irrelevant chatter.
TL Za 1 presents a straight-forward statement of the first part of the Formula:
In a more complicated variation on the Formula, ZA Zb 3, we can infer some of the same types of words:
In this statement, we see the Verb A-TA-I-*301-DE-KA farther along.
The word order here must therefore start, not with Verb, but with the Object, "fronted for emphasis" (p. 378). After this 3-word sequence for the Object we find the Verb A-TA-I-*301-DE-KA (past tense?) foillowed by the expected Subject A-RE-PI-RE-NA TI-TI-KU.
The 3-word sequence between the Object and the Verb should be the recipient, the Indirect Object DI-DI-KA-SE A-SA-MU-NE A-SE.
Units (the number 'one') are invariably written as vertical strokes, up to nine of these.
Tens are usually written as horizontal strokes (again up to nine of these) but at Knossos, Malia, and Phaistos, tens are written as simple dots.
Hundreds are conveyed by circles.
Thousands are conveyed by circles with rays, usually four, each from the cardinal points, e.g., HT 31, but MA 9 seems to use circles with random spikes around the circumference: .
HT 31 employs
all these conventional numeral signs:
|JJ (PH 9b)|
|J||A (HT 120.3)|
|J||E (lots; EJ [HT 123a.3-4; ZA 8.4])|
|J||E||B (HT 27a.8)|
|J||B (HT 129.1; KH 5.4, 6.8, 17.3)|
|J||F (HT 51b.2)|
|J||K (HT 32.1)|
|J||H (HT 93a.3)|
|J||E||L2 (KH 7a.5, 56.1)|
|J||L2 (HT 123b.4)|
|EE (PH 12b.2, 13a, 13c)|
|E||B (KH 9.2)|
|E||F (HT 8b.4, 16.3, 40a.4, 123b.5, Zd 156)|
|E||L2 (HT 33.3)|
|E||L4 (KH 26.2)|
|E||L6 (KH 76.2)|
|E||YYY (PH 26)|
|A||BB (KH 86.2)|
|BB (KE Wc 2b)|
|F||K (PH 1b.2)|
|F||L (ZA 7b.8)|
|H||K (HT 34.3)|
|K||L2 (HT 86a.2, 120.2; KH 11.2.3-4, 4.6, 16.1, 75.2)|
|L2L4 (HT 33.2)|
|L3L3 (HT 15.2)|
If, however, X (A711) is the half-mina and W is the mina, then W would
be 1/60 talent -- that might explain the formal relationship between B (if
B were 1/3) and
W: W would be a conjoined BB with an implied value of 1/10 x 1/6.
Or it may be the Linear A equivalent of Linear B *116 N, the half mina -- if so, then W could be the full mina (see above).
For four of the fractions (J, E, F, K), we can demonstrate certain values, and can suggest, with much less certainty, values for an additional five fractions. The chart below also gives Hiero fractions that are similar in shape.
The horizontal lines below the fractions point out "family" resemblances in shape.
|Linear A||Linear B||Denomination||Mass (gr, approximate)||Fraction of Talent||Fraction of preceding|
|A717, DD||B117, M||double mina||967||1/30||1/30|
|A711, X||B116, N||1/2 mina||242||1/120||1/2|
|-||B115, P||1/24 mina||20.2||1/1440||1/12|
|-||B*21, Q||1/144 mina?||(3.36)||1/8640||1/6|
In texts that employ a string of signgroups, dots separate them. This practice is most notable on non-bureaucratic texts (like the hair pins, e.g., CR(?) Zf 1) and especially in dedicatory texts like those on the "Libation Tables" (e.g., IO Za 2).
|.1-2||RA-*164a-TI|| TE VINa||30|
The documents that employ the Continuity Principle might therefore reflect short initial documents that record contributions which are then organized in an outline fashion. For instance, I can imagine 5 separate documents being collated to produce HT 20:
The implication of such a system is that the 5 separate short texts (really more like chits) were brought together, not because they represent the same contributions or contributions from the same place/person, but because they represent contributions organized according to a larger principle: 1) made at the same time, or 2) made from the same region/person, or 3) organized/collected by the same Collector.
Comments, corrections, questions: John