Balashon’s latest post, bareket and emerald, is about a connection I had forgotten:

From Hebrew (or some other cognate Semitic language, like the Akkadian barraqtu), bareket entered into Greek as smaragdos, which Latin borrowed as smaragdus, eventually becoming esmaraldus in Medieval Latin, esmeraude in French, and then “emerald” in English.

This might seem like a strange journey, particularly from bareket to smaragdos. But as this Philologos column explains (along with many other interesting linguistic details about the words we’ve discussed here and more) it’s reasonable when you look at how certain letters are exchanged in phonetic shifts.

Philologos actually promotes a different theory than what I’ve presented here. He says that the Hebrew baraket may have its origin in a Sanskrit word – marakata […]

Most of the sources I looked at, including Klein and the Online Etymology Dictionary say the Sanskrit word was borrowed from a Semitic source. (For further discussion see this page).

I say “had forgotten” because it turns out I wrote about it in 2004:

I enjoyed [Philologos’s] detailed investigation of the etymology of Yiddish shmergl ‘emery,’ which traces it back to Latin smericulum and Greek smaragdos ‘emerald’; I think the bald assertion that the latter is borrowed from Sanskrit marakata goes beyond the evidence, but this is, after all, a newspaper column, not a linguistic journal.

AHD fudges the details of the relationships with not one but two instances of “akin to”:

[Middle English emeraude, from Old French, from Medieval Latin esmeralda, esmeraldus, from Latin smaragdus, from Greek smaragdos; akin to Sanskrit marakatam, probably of Semitic origin; akin to Akkadian barraqtu and Hebrew bāreqet, a kind of gemstone (probably emerald); see brq in the Appendix of Semitic roots.]

Anybody know anything more about this tangle?


  1. To complete the cycle, you will also find Hebrew use of אזמרגד, /izmargad/, a re-loan of the Greek form.

  2. The following is from Benjamin Noonan’s 2012 dissertation, Foreign Loanwords and Kulturwörter in Northwest Semitic (1400-600 B.C.E.): Linguistic and Cultural Contact in Light of Terminology for Realia (it’s now been revised as a book but I don’t have it.)

    This word occurs only three times in the Hebrew Bible, twice in the form בָּרֶ֫קֶת (Exod 28:17; 39:10) and once in the form בָּֽרְקַת (Ezek 28:13). [305] In Exod 28:17-20; 39:10-13, it appears within the description of the high priest’s breastplate along with several gemstone terms that are foreign loans. Speaking of the king of Tyre’s adornment, Ezek 28:13 mentions בָּֽרְקַת in conjunction with many of these same gemstones. [306]

    Köhler and Baumgartner postulate that Hebrew בָּרֶ֫קֶת is derived from Sanskrit marakata, “emerald.” [308] However, Sanskrit marakata is a late loan from Greek σμάραγδος (in turn recognized by classicists as a loan into Greek from Semitic) [309] and thus cannot be the origin of the Semitic forms. [310] Hebrew בָּרֶ֫קֶת is based on the common Semitic root brq, “to flash, shine,” [311] an appropriate etymology for a gemstone. Attested cognates exist in Syriac (bārqā) [312] as well as Neo-Babylonian Akkadian (barraqtu). [313] Based on its lexical derivation as well as the testimony of classical sources, בָּרֶ֫קֶת and its Semitic cognates could designate a variety of shiny gemstones, including (but not limited to) the emerald. [314]

    Some footnotes:

    [309] LSJ 1080, 1619; EDG 1365-1366; DELG 991, 1353. Mayrhofer and Beekes plausibly suggest that the spelling of σμάραγδος represents an attempt by the Greeks to folk etymologize this term on the basis of the verb σμαραγέω, “to thunder” (LSJ 1619); see EDG 1365-1366; Manfred Mayrhofer, “Indogermanistische Randglossen zu ‘Kluge-Mitzka,’” Die Sprache 7 (1955): 187-188. The form μάραγδος does not occur in Greek until late, namely the first century BCE, and may be a reborrowing from Sanskrit (EDG 1366).

    [310] KEWA 2:587-588; Mayrhofer, “Indogermanistische Randglossen zu ‘Kluge-Mitzka,’” 187-188. Egyptian brgt, attested only once in the Late Period Sehel Inscription (line 16), is a loan from Semitic in light of its late attestation and rarity in Egyptian (Harris, Lexicographical Studies in Ancient Egyptian Minerals, 105). Pahlavi uzumburd and New Persian zumurrod as well as their derivative Arabic zummurud, in turn, are loans from Greek σμάραγδος as demonstrated by their representation of an initial sibilant (Lane 1251; CPD 85; CPED 621; NPED 1:1018; Asbaghi, Persische Lehnwörter im Arabischen, 147).

    [314] Classical descriptions of σμάραγδος and smaragdus (e.g., Theophrastus, De Lapidibus 4.23-27; Pliny, Nat. 37.16-19; Diodorus Siculus, Bibliotheca historica 2.52) indicate that a variety of different gems were denoted by this term in antiquity, but not exclusively the emerald; see John Sinkankas, Emerald and Other Beryls (Radnor, Pa.: Chilton Book Company, 1981), 13-21; John F. Healy, Pliny the Elder on Science and Technology (Oxford: Oxford University Press, 1999), 241-245. Probably due to its hardness, usage of the emerald is not attested in Mesopotamia prior to the fourth century BCE and in Egypt prior to the Ptolemaic period (Moorey, Ancient Mesopotamian Materials and Industries, 81; Aston, Harrell, and Shaw, “Stone,” 25).

  3. Thanks very much!

  4. Chantraine is skeptical: “Le σ- initial du grec n’est pas expliqué, une influence de σμαραγέω n’est guère plausible.” Beekes adds: “For the Greek reflex σμ-, cf. Σμέρδις beside OP Bardiya, etc.” I don’t know what’s behind the “etc.”; borrowing a foreign [b-] as [sm-] is obviously weird, but apparently the Greeks did it more than once.

  5. Another of the mystery minerals in the Egyptian breastplate is פִּטְדָה piṭdā. There is general agreement on its identification as peridot, which came from an island in the Red Sea. It is likely a loan (odd phonotactics, no native etymology), and a Nubian origin was suggested, a guess based on no more than geographical location. I don’t know if anyone who studies Nubian languages has looked into it.
    The word peridot itself, which first appears in Europe in the Middle Ages, is of “uncertain etymology” per the OED. I wonder if there is some connection between peridot and piṭdā ~ piṭǝdā.

  6. Henry and Renée Kahane offer an interesting etymology of the word peridot in their article “Romano–Aegyptiaca” (Romance Philology 14, no. 4 (May, 1961), pp. 287-294):


    (JSTOR offers 100 articles a month with free registration for LH readers who do not otherwise have access to this publication.)

  7. They derive it from “the Greek designation of a kind of opal or amethyst, paidéros (acc. paidérōta), first recorded by Pliny.”

  8. “borrowing a foreign [b-] as [sm-] is obviously weird”

    It brings to mind the recent re-transliterations of Burma as Myanmar and Bombay as Mumbai. Apparently there is a consonant that can be heard as b in one language and m in another. (I know, that doesn’t account for the s.)

  9. It brings to mind the recent re-transliterations of Burma as Myanmar and Bombay as Mumbai. Apparently there is a consonant that can be heard as b in one language and m in another.

    I don’t think this is what happened; “Burma” comes from Bamar, the common name of the largest ethnic group in the country – and “Myanma” is a more formal alternative name for the Bamar. “Mumbai” is after the city’s patron goddess, Mumba, but “Bombay” is I think Portuguese in origin, “good bay” – the brownshorts argued that Bombay is an English corruption of an original name Mumbai, but they may not be the best sources.

  10. David Marjanović says

    Classical Burmese myan has been reduced to a toneless [bə] in the modern language, and the r is a British orthographic device to indicate vowel length. Shifts of nasal consonants to voiced plosives are not common worldwide, but the Min “dialects of Chinese”, including the one called Taiwanese, have shifted the whole series of syllable-initial [m n ŋ] to [b d g] across the board.

  11. “Brownshorts”. Thanks for that.

Speak Your Mind