Minoan and Mycenaean fig trees: some retrospective and prospective comments

2019.12.27 | By Gregory Nagy (corrected 2020.01.02)

§0. The photograph I have chosen to illustrate my all-too-brief comments for this posting shows a fig tree. My first impression when I look at any adult fig tree in general is that its branches seem to crisscross each other, as we can see in the photograph I show here. And this visual pattern of crisscrossing reminds me of the ideogram representing this tree in the scripts used by Minoan and Mycenaean scribes in the second half of the second millennium BCE. In both scripts, Minoan Linear A and Mycenaean Linear B, the ideogram is shaped like the capital letter Y, but each one of the two branches of this Y is crisscrossed by one counter-branch, so that the vertical stem of the tree, shaped like a capital I, gets crowned by two sets of branches, and each set is shaped like a capital X. In this posting, I will show what I think are two examples of this ideogram, written in Linear A. The writing is on the surface of two clay sealings found at the site of the Minoan palace at Hagia Triada in Crete, dating back to the fifteenth century BCE. These sealings, technically “roundels,” were receipts for commodities. (I have corrected my original wording here, following the advice of John G. Younger about the function of “roundels.”) At the center of both these sealings, in terms of my argument, is the Linear A ideogram for {FIG TREE}, and, at each side of the ideogram, are two signs for two syllables that spell, if you put them together, a word that means ‘figs’ in Greek as spoken in the post-Minoan-Mycenaean era of the first millennium BCE.

Fig tree. Image via Flickr, under a CC BY-NC-ND 2.0 license.
Unicode version of a Linear B sign standing for the syllable -ni-, corresponding to a Linear A sign with the same shape and standing for a comparable syllable, arguably also -ni-. The shape of this sign matches the shape of an ideogram, used in both the Linear A and the Linear B scripts, that stands for {FIGS}; this ideogram is a visual representation of a fig tree.

§1. The argument I have just outlined was already formulated in two articles I published more than half a century ago, Nagy 1963:198–199 and Nagy 1965:321. Since those publications, however, I have not returned to my study of Linear A, but now, as I look ahead to future projects, I hope to update my views in the light of research ongoing over the last five decades. In the course of this anticipated updating, I will be revising my references to the corpus of Linear A texts, which have by now been thoroughly re-edited (Godart and Olivier 1976–1985). With these prospective remarks in place, however, I will for the moment refer retrospectively to the two sealings, as described at §0 above, in terms of my original presentation in Nagy 1963:198–199, where I was using an older edition of the Linear A corpus (Brice 1961). In that edition, these sealings are classified as Cr IV 2a and Cr IV 3a:

Two sealings from the palace at Hagia Triada, Crete: “Cr IV 2a” and “Cr IV 3a” (in the edition of Brice 1961). Facsimiles (after Brice) from Nagy 1963:198.

§2. In Nagy 1963:198–199, I offered an analysis that I rewrite here, making changes in format but not in content (except for additional comments, enclosed within square brackets):

The three figures written on the surface of both “Cr IV 2a” and “Cr IV 3a” are interpreted as one word in the edition of Brice 1961, and [in terms of the format generally used nowadays for transcribing Linear A] such a word would spell “SU-NI-KA.” [Though the first figure in “Cr IV 3a” looks more like a “TA” than a “SU,” the reading “SU” is preferred  for both “words” in the edition of Brice 1961.] But I proposed an alternative reading: since the position of the figure in the middle, transcribed as NI, is a little higher than the positions of the two enclosing figures transcribed as su- on the left side and -ka on the right side, I chose to read the highlighted figure in the middle not as the syllable -ni- but as the ideogram used in the Linear A script to designate {FIGS}. And then the two enclosing figures, transcribed as the syllables su- on the left side and -ka on the right side, could be read as su-ka, with the ideogram {FIGS} superimposed. Such a word, written this way in the second millennium BCE, could be compared with a common word used in the Greek language as attested in the first millennium BCE, which is sūka in the plural / sūkon in the singular, meaning ‘figs/fig’. This word for ‘figs’ has not yet been found in Linear B texts, but the derived word for ‘fig tree’ is in fact attested as su-za, from sūkia.

§3. The Linear B form su-za is attested in Knossos tablet F 841+867 line 5, followed by the ideogram NI = {FIGS} plus a measure amounting to 75, followed—after some spacing—by the word ka-po = karpos ‘harvest’. I add here an observation that is relevant to my proposed reading of signs indicating the commodity of ‘figs’ as written in Linear A script on sealings. As I observed in Nagy 1963:198 (also at p. 187), we find written in Linear A script on a roundel classified as Cr IV 51 (in the edition of Brice 1961) three signs that can be read syllabically as ku-mi-na. These signs, as I argued, indicate the commodity of ‘cumin’. Linear A ku-mi-na is comparable to Linear B ku-mi-na, meaning ‘cumin’, which is found in the text of Mycenae tablet Ge 605 both as plural ku-mi-na and as singular ku-mi-no, to be transcribed as kumīna and kumīnon, written as κύμινα and κύμινον in classical Greek. The word is non-Indo-European, well-attested also in Semitic languages.

§4. The analysis that I have offered so far, I must emphasize, depends on a basic fact about Linear A and Linear B: the fact is, Linear B actually inherited from Linear A the practice of writing ideograms and ligatured logograms. And my analysis depends also on another basic fact—at least, it is a fact in my opinion. The fact is, the Linear B script also inherited its syllabary from Linear A. As I showed in my two articles Nagy 1963 and 1965, and as many others have shown since then—here I single out the exemplary work of Packard 1974—the figures used for representing syllables in Linear B correspond phonetically to the figures used for representing syllables in Linear A. The correspondences may not be exact, and there is much ongoing debate about the relative degrees of exactness, but it cannot be denied that as many as fifty of the syllabic signs in Linear A are matched to some degree by fifty corresponding signs in Linear B. And among these signs in Linear A is what I transcribe as -ni-, which looks exactly like the ideogram for ‘figs’ in Linear A—and in Linear B.

§5. That said, I return to my argument in Nagy 1963:198–199 concerning what is written on two sealings found in the palace of Hagia Triada in Crete, dating from the middle of the second millennium BCE. As I argued, the Linear A writing on these sealings can be read as a combination of the ideogram meaning {FIGS} with two syllables su- and -ka that spell the word sūka meaning, likewise, ‘figs’. What I should have added to my argument published in 1963, and I add now, is a relevant fact: the Linear A and Linear B ideogram for {FIGS}, which is known to have the syllabic value of -ni- in Linear B, has been explained by Günter Neumann (1962) as an “abbreviation” of nikúlea, which was a Cretan word for ‘figs’, as we know from Athenaeus (3.76e), mediating a work by Hermonax on Cretan vocabulary: Ἑρμῶναξ δ’ ἐν Γλώτταις Κρητικαῖς σύκων γένη ἀναγράφει ἁμάδεα καὶ νικύλεα ‘Hermonax in his treatise “Cretan glosses” notes two kinds of figs [sūka]: they are amádea and nikúlea’. Such an “abbreviation” can be seen as acrophonic. In other words, just as we have “a as in apple” in alphabetic writing, we have ni- as in nikúlea (= νικύλεα) ‘figs’ in the syllabic writing of Linear A and Linear B, where the acrophonic abbreviation functions also as an ideogram (Thompson 2012:552–553, with reference to Ruijgh 1962:32; also Duhoux 1997:71).

§6. I should add that the evidence of the Linear A texts found at Hagia Triada points to more than one word that was used to designate ‘figs’. In the tablet HT 88, line 2, the ideogram {FIGS} is followed by three figures that together spell syllabically the word ki-ki-na, followed by the number 7 indicating the measured quantity of this commodity. Here I turn to another article by Günter Neumann (1960), showing evidence for the survival of this word into the first millennium BCE, where we see it attested in the dictionary attributed to Hesychius: κεικύνη· συκάμινος. The form keikunē, which most likely reflects an earlier *kikunā, is glοssed here as sukaminos, which can refer either to a mulberry tree or to a kind of fig tree (as we read for example in Diodorus of Sicily 1.34.8). In this connection, I find it relevant to quote here a comment by John Younger (2019.09.08), who considers the possibility that two different terms were used in the Linear A texts in referring to two different kinds of figs: “Perhaps the two terms characterize ‘fresh’ and ‘dried’ figs or ‘green’ and ‘black’ figs.”

§7. So, does all this mean that the Linear A script can be read as “Greek”? I had no answer to that question back when I published my views about Linear A in Nagy 1963 and Nagy 1965, and I still have no answer. But I am ready to add now an idea that I had not yet fully formulated back then: there is no need to suppose that the Linear A script was designed to represent one single language that archaeologists today might be tempted to describe as “Minoan.” Yes, the “Mycenaean” language represented by Linear B was Greek, but I doubt that Linear A represented any kind of unified “Minoan” language. And I cannot rule out the possibility that the various different writers and readers who had access to the Linear A script may have included at least some speakers of the Greek language.


Brice, W. C., ed. 1961. Inscriptions in the Minoan Linear Script of Class A. Oxford.

Duhoux, Y. 1997. “Le linéaire A: problèmes de déchiffrement.” In Problems in Decipherment, ed. Y. Duhoux, T. G. Palaima, and J. Bennet, 59–119. Louvain-la-Neuve.

Godart, L., and J.-P. Olivier, J. P., eds. 1976–1985. Recueil des inscriptions en Linéaire A [abbreviated GORILA]. Études Crétoises 21, vols. 1–5. Paris.

Nagy, G. 1963. “Greek-like elements in Linear A.” Greek, Roman, and Byzantine Studies 4:181–211.

Nagy, G. 1965. “Observations on the Sign-Grouping and Vocabulary of Linear A.” American Journal of Archaeology. 69:295–330.

Neumann, G. 1960. “Minoisch ki-ki-na ‘die Sykomorenfeige’.” Glotta 38:181–6.

Neumann, G. 1962. “νικύλεον.” Glotta 40:51–54.

Packard, D. W. 1974. Minoan Linear A. Berkeley and Los Angeles.

Ruijgh, C. J. 1962 Études sur la grammaire et le vocabulaire du grec mycénien. Amsterdam.

Thompson, R. J. E. 2012. “In Defence of Ideograms.” Études Mycéniennes 2010: Actes du XIIIe Colloque International sur les Textes Égéens, ed. P. Carlier, C. de Lamberterie, M. Egetmeyer, N. Guilleux, F. Rougemon, J. Zurbach, 545–561. Pisa/Rome.

Younger, J., ed., with commentary, 2019.09.08. Linear A Texts in phonetic transcription. Online edition, http://www.people.ku.edu/~jyounger/LinearA/HTtexts.html.