Some Notes on the
History of Isopsephia (Gematria)

John Opsopaus

Georges Contenau (p. 166) cites evidence that isopsephia (the Greek word for gematria, which he uses) was widely used in Mesopotamia. In this case numerical values were assigned to the characters in their syllabary, and the numerical values of names were computed. (This also fits in with the general Mesopotamian reverence for numbers: all the Gods had numbers, which Simo Parpola has shown to be numerically related by the Assyrian Tree of Life.) For example, Sargon (d.705 BCE) states that the perimeter of his palace at Khorsabad (16283 cubits) was equal to his name.

The next datable source for gematria I know about comes from David Fideler's Jesus Christ: Sun of God. He argues that the standard spellings of the Gods' names were formulated according to isopsephic principles under the influence of the Pythagorean League c. 500 BCE (p. 75). So for example, Zeus is the Geometric Mean of Hermes and Apollo (p. 72). He further argues (pp. 216-9) that many Greek temples, such as the Parthenon (447 BCE) and Apollo's temple at Didyma (300 BCE), were constructed isopsephically.

This use of gematria in temples and other official buildings agrees with its only known etymology, from Greek geometria (earth-measures).

Given the early use of gematria in Babylonia (by the eigth century) and its apparent use in Greece (by the fifth), the general idea must have been widely known around the Mediterranean from an early date.

Nevertheless it seems that the Hebrew gematria known to us now (i.e., assigning 1-9, 10-90, 100-900 to the letters) came from the Greeks. (Of course there are other ways of doing gematria; sometimes the Greeks assigned 1-24 to their letters.)

The best discussion I've found so far on the historical issues and evidence is in Ifrah (chs. 16 & 17). He says (p. 267) that the Greek numerals go back "at least to the end of the fourth century B.C."; they appear on coins (266 BCE) and papyri (311 BCE). However, "the oldest examples of the Hebrew system go back only to the beginning of the first century B.C. or, at most, to the last few years of the second century" (p. 267). In particular, the oldest example is a coin dated to 103 BCE (p. 269). Prior to that the Jews used other numeral systems (Egyptian hieratic during the monarchic period, Assyro-Babylonian sexagesimal later, etc.; see p. 270). He adds (270),

These considerations support the view that the Hebrew alphabetic numeration was derived from the Greek, and this view is certainly not contradicted by Jewish inscriptions made later, during the Diaspora. Between the first century B.C. and the seventh century A.D., when use of the Hebrew alphabetic numeration was becoming increasingly common in the Jewish world, several Jewish scribes in the Mediterranean basin (from Italy to northern Syria and from Phrygia to Egypt), who could write as well in Hebrew as in Greek or Latin, often continued to use the Greek numeral letters.
(See also Menninger, pp. 264-5, who argues for a Greek origin.)

An additional argument for the Greek origination of this system is that the Greek archaic alphabet had the 27 letters necessary for the numbers (1-9, 10-90, 100-900), whereas all the Northern Semitic alphabets at that time had 22 letters; numbers greater than 400 were expressed by combining taw with other letters (qoph, resh, shin, taw, additively). Thus, for example, taw taw = 800 (Cajori, v. I, pp. 19-20).

The familiar system of gematria requires 27 letters for three enneads (1-9, 10-90, 100-900); therefore Hebrew gematria depends on the assignment of separate numerical values to the medial and final forms of kaph, mem, nun, pe and sade.

According to Diringer (Story Aleph Beth, 136), one of the distinctive differences between Early Hebrew (which is very much like Phoenician) and Square Hebrew (which is influenced by and perhaps derived from Aramaic) is the presence of final forms in the latter, so this form of gematria cannot have been used before Square Hebrew. Diringer (135) says, "a distinctive Palestinian Jewish type of script - which we can definitely regard as the Square Hebrew script - can be traced from the second and the first centuries B.C.E. According to Prof. W. F. Albright, it became standardized just before the Christian era." He adds (136), "In the Square Hebrew style, unlike the Early Hebrew, ... there are five letters which have dual forms, one when initial or medial, the other when final.... The dual forms in great part go back to the period before the various offshoots of the Aramaic script assumed their distinctive features; they are found, indeed, in some third-century B.C.E. cursive documents in Egypt, in Nabataean inscriptions, and in the earliest Square Hebrew inscriptions and other documents. In some early documents, the letters 'aleph, he and taw also have dual forms." And finally (p. 137), "It was during the second century C.E. - according to Prof. Tur-Sinai - that our present Square Hebrew script, in its current form, became more or less fixed, and it was only in this period that the consistent Massoretic tradition regarding the use of the dual forms of the letters kaph, mem, nun, pe, sade was established."

Therefore the use of the final forms in gematria is not likely to predate the second cent. C.E., and their use in numeration cannot have predated their invention in the third or second cent. B.C.E. (Further, as noted above, they were originally assigned the same numerical values as the corresponding medial forms, which would agree with their originally inconsistent use in medial and final position.)

I have also found some relevant information on the history of the Greek alphabet. There is considerable evidence that the Greek alphabet was derived from a proto-Canaanite script before 1050 BCE (Naveh 178). In fact this is the approximate chronology given by Naveh (10):

BCE 1700 proto-Canaanite developed 1100 archaic Greek derived from proto-Canaanite 1050 proto-Canaanite -> Phoenician 800 (Early) Hebrew derived from Phoenician 700 Aramaic derived from Phoenician 650 Latin derived from Greek 200 Square Hebrew (or Jewish) derived from Aramaic This is important because the Archaic Greek alphabet had the 27 letters needed to support the three enneads of a complete number system, and it was used this way throughout antiquity (even after three of the letters were no longer used for ordinary writing).

Martin Bernal goes further. He argues that between 1750 and 1400 BCE the Greeks borrowed an early Semitic alphabet with 28 consonants. (Other authors have argued that they correspond to the Mansions of the Moon - but that's another story!) The order of the letters was that documented for the (long) Ugaritic alphabet, which also derives from this Semitic alphabet. Later Semitic alphabets reduced the number of consonants to the 22 familiar ones, but archaic Greek kept 27 of the original letters, though seven of them were reallocated to vowels (as indeed Semitic writers had sometimes used them). Much later, early in the first millenium BCE, under influence from the (22 letter) Phoenician alphabet, the Greeks reorganized their alphabet and borrowed the Phoenician names of those 22 letters (aleph -> alpha, etc.), and their order (non-Phoenician letters were moved to the end). At this time the alphabetic numeral system (as used in gematria/isopsephia) was also established (late tenth or ninth cent. BCE). Bernal adds (125), "This fits well with the historical and archeological evidence for close contact between the Levant and the Aegean in this period." Perhaps we should not be surprised that this chronology agrees with Herodotus' claim that the alphabet was brought to Greece in the Bronze Age (mid-second millenium BCE) by "Phoenician" (i.e. Levantine) colonists. (See Bernal chs. 3, 5, 6, esp. pp. 61, 66-9, 114, 123-8)

In summary, the early borrowing by the Greeks of a 28-letter Semitic alphabet is consistent with the (numismatic and papryrological) evidence for the development of the three-ennead alphabetic numeration system in Greece (possibly by the tenth cent. BCE, but definitely by the fourth cent. BCE).

Of course this evidence doesn't preclude the possibility that the Jews already knew about gematria from the Babylonians, or that they were doing it on some other basis (e.g. assigning 1-22 to the letters). Certainly new data could turn up to reverse the order, but that's always the way it is with history!


  1. Bernal, Martin. Cadmean Letters: The Transmission of the Alphabet to the Aegean and Further West before 1400 B.C. Winona Lake: Eisenbrauns, 1990.
  2. Cajori, Florian. A History of Mathematical Notations (2 vv.). Chicago: Open Court, 1928-9.
  3. Contenau, Georges. Everyday Life in Babylon and Assyria, New York: St. Martin's Press, 1954.
  4. Diringer, David. The Alphabet: A Key to the History of Mankind (2 vv.). New York, Funk & Wagnalls, 1968.
  5. Diringer, David. The Story of the Aleph Beth. New York: Thomas Yoseloff, 1960; World Jewish Congress, 1958.
  6. Fideler, David. Jesus Christ: Sun of God: Ancient Cosmology and Early Christian Symbolism. Quest Books, 1993.
  7. Ifrah, Georges. From One to Zero: A Universal History of Numbers, Viking, 1985.
  8. Menninger, Karl. Number Words and Number Symbols: A Cultural History of Numbers. Cambridge: MIT Press, 1969.
  9. Naveh, Joseph. Early History of the Alphabet: An Introduction to West Semitic Epigraphy and Paleography. Jerusalem: The Magnes Press, The Hebrew University, 1987.
  10. Parpola, Simo. "The Assyrian Tree of Life: Tracing the Origins of Jewish Monotheism and Greek Philosophy." Journ. Near Eastern Studies 52, 3 (July 1993), 161-208.

Return to Biblioteca Arcana page

Send comments about this page
Last updated: Tue Jun 17 08:53:06 EDT 1997