God in Four Letters.

I do love a good crackpot — excuse me, I mean premodern — etymology, and Poemas del río Wang has a doozy; the quote is from “the eleven-language dictionary of Calepinus, published in 1590 in Basle, which I happen to have here on my bookshelf” (“I found this bulky folio volume some thirty year ago in a waste paper recycling shop, and purchased it at the price of scrap paper, for about one euro in today’s currency”):

It is not unworthy to consider, that almost every people and language writes the name of God in four letters. In fact, the Hebrews call him יחוח Yehova, with four letters, the Chaldeans also with four letters, אלוח Eloha, the Syriacs also אלוח Eloha; at the Aethiopians He is אמלו Amlau, at the Assyrians אדעד Adad, at the Greeks Θεός, at the Egyptians Θωύθ, at the Persians Σύρη, at the Latins Deus, at the Italians Idio, at the Spanish Dios, at the French Dieu, at the Germans, Flemish and English Gott or Godt, at the [Persian] Magi Orsi, at the Poles Boog, from bog, that is, ʻfear’, at the Dalmatians and Illyrians Boga or Boog, at the older Muslims, whom we also call Saracens, Abgd, at the Turks following Mohamed Alla, at the peoples discovered in the world called “new” Zimi, at the Vlachs Zëul, at the Gypsies Odel.

At the Hungarians, if we look at its origin, the name of God has also four letters. They call him with great respect Isten, which, although seems to have five letters, if we consider its origin, has only four. In fact, the Hungarian term comes from the second aorist of the verb ʻto be’ ἴστημι, which sounds ἐϛὶν [ἐστὶν]: ʻI exist, I am by way of myself’, which second aorist is written with four letters. The s and t, written with two letters in the Hungarian word, are both encompassed in the single Greek letter ϛ sigmatau. Thus, by virtue of its origin, the Hungarian name also has to be written with four letters, so: Ἴϛεν [Ἴστεν]. In this way, the name of God is a τετραγράμματον [four-letter name] for every people, and we think He is called so, because His essence is one, but within His one essence He is three actually existing and different persons.”

You can see an image of the Calepinus page at the link, where you will also find a discussion of the actual etymology of Hungarian isten (about which there are various opinions).


  1. This must be the source of Athanasius Kircher’s better-known diagram to this effect.

  2. In all three Hebrew-character forms you (well, he) replaced the four ה-s with ח-s.

  3. Very entertaining stuff. But Calepinus is slightly wronged by the translator: ἴστημι means “stand”, as he says (“subsisto, per me ipsum sto”), not “be”. He gets the aorist form wrong (it should be ἔστην), presumably misled by the iotacistic modern pronunciation.

  4. J. W. Brewer says

    The odd thing is that the full entry scanned/transcribed on at the banks of the Wang gives the standard three-letter form for English yet the commentary lumps the English in with the German/Dutch as using a four-letter form. Maybe in some earlier version of English? Although the first searchable allegedly-original-spelling Chaucer I googled up seems to use “God” consistently for the simple form although there is a letter-doubling to “Goddes” when it’s possessive.

  5. Now we have discovered the true meaning of the Russian orthographic reform of 1918. It was the Bolshevik design to change Богъ to Бог and ruin the Trinity.

  6. The OED s.v. god says (in full, but without markup):

    α. eOE go (Mercian, transmission error), eOE gotho (plural, rare), OE geode (dative, transmission error), OE geodes (genitive, transmission error), OE godedes (genitive, transmission error), OE (rare) ME–16 godd, OE–eME cod (rare), OE–eME goð (perh. transmission error), OE– god, lOE gedes (genitive, transmission error), lOE gode- (in compounds), lOE goden (dative, perh. transmission error), eME ȝod, eME ȝodd, eME got, eME goðð (perh. transmission error), ME godde, lME gottys (genitive), 15 gos (genitive); also Sc. pre-17 godd; chiefly N. Amer. (colloq. and regional) 18– gahd, 19– gaahd.

    β. OE (rare) ME–15 good, eME ȝode, ME goed, ME goid, ME gooddes (plural), ME guodes (genitive, perh. transmission error), ME–16 gode, lME goodyse (genitive), lME goyd; Sc. pre-17 goid, pre-17 19– gode, 18– goad.

    So four-letter forms (presumably to show the short vowel) have existed in English, but are rare, and many of them are typos (“transmission errors”). Note the 18th-century beginnings of the “father-bother” merger in AmE in the spelling “gahd”, which faithfully represents my pronunciation, though “gawd” is not unknown.

    There is a story about a child whose family moved from the Midwest to Boston, where the kid heard talk on the schoolyard of someone who he interpreted as a legendary giant. When his parents wanted to know the name of this giant, he translated it from eastern New England to Midwestern and said “They call him Guard.”

  7. Huh, the OED entry was updated just this June. Here’s their new etymology:

    Cognate with Old Frisian god, Old Dutch god (Middle Dutch, Dutch god), Old Saxon god (Middle Low German got, (inflected) gōd-, godd-), Old High German got (Middle High German got, German Gott), Old Icelandic guð, goð, Old Swedish guþ (Swedish gud), Danish gud (already in early modern Danish), Gothic guþ (usually abbreviated as ḡþ) < a Germanic base of uncertain origin …

    The further etymology is very uncertain. The underlying Germanic base has been explained as a derivative (with the Germanic base of –ed suffix 1) of the zero-grade of either of two possible Indo-European verbal bases:

    (i) the (unsuffixed) base of yet v. with the underlying meaning ‘pour’ (used here with reference to poured sacrifices or libations), or (ii) a base with the meaning ‘to invoke’.

    With (i) perhaps compare Sanskrit huta poured in sacrifice, sacrificed and the similarly-formed (or perhaps cognate) ancient Greek χυτός poured (out), fluid.

    The base of (ii) is apparently reflected by Sanskrit – to invoke (a god) (ablaut variant of hve– to call, summon; compare also puru-hūta, lit. ‘much-invoked’ or ‘invoked by many’, epithet of Indra), Old Irish guth voice, Old Church Slavonic zŭvati to call out, invite, name; this interpretation has often been favoured, but it poses phonological problems.

    Compare also Gaulish gutu– (in gutuater priest), Lithuanian žavėti to charm, practise magic, which may derive in a similar way from either of these Indo-European bases.

    Note that the Germanic forms are almost all triliteral, though in Gothic they managed to boil it down to two letters and a diacritic.

  8. J. W. Brewer says

    Hmm. Calepinus gives “Godt” as the “Belg.,” word, by which I assume he means Dutch. I have actually read the Gospel in Dutch aloud in public at polyglot Easter services (despite not actually knowing Dutch – I just treated it as an oddly-spelled German dialect and tried my best to read it with the correctly-odd accent) but did not commit spellings to memory, so “Godt” didn’t immediately strike me as Obviously Wrong. John 1:1 in the classic 17th century translation is “In den beginne was het Woord, en het Woord was bij God, en het Woord was God.” However, given that readily-accessible editions of the King James Version almost always have modernized spelling with substantial variants from the 1611 original, I’m not sure that pulling up online the text I just quoted is direct evidence of actual 16th/17th C. Dutch orthography.

  9. Stefan Holm says

    A maybe peculiar fact is that the word ‘god’ originally wasn’t ‘he’ or ‘she’ but ‘it’, neuter, in Gmc. In Gothic it lacks the masculine ‘-s’ marker in nominative and in ON the corresponding ‘-(e)r’ marker. C.f. Christ who in the oldest Swedish source is Krister.

  10. Just for the record: the Polish word is Bóg, not “Boog”, and bog doesn’t mean ‘fear’ in Polish (as a matter of fact, there’s no such word). The older meaning of Slavic *bogъ (still detectable in compounds and derivatives) was ‘wealth’. It’s in all likelihood an Iranian loan, and the religious extension of the meaning is also parallel to what we find in the Iranian languages.

  11. Is there an accepted explanation of how Icelandic Guð came to be pronounced [gvʏð]?

  12. People seem to have been fascinated by this idea for a long time. An 1882 magazine of leisurely reading received the inquiry, “I recently saw an item in some newspaper that the name of God is spelled in nearly all languages with four letters. Is there any explanation of it? Can some one furnish a list of languages and spellings of the name?—Z.” The magazine provides the cowardly one-lettered correspondent with a list of purported such words.

    I’m not sure what language ‘zimi’ is supposed to be in. This list says it’s in the language of the “Peruvians”. Mangled Quechua?

  13. Stefan Holm says

    how Icelandic Guð came to be pronounced [gvʏð]?

    I’m not sure it did. Standard modern Icelandic pronunciation of ‘u’ is IPA /ʏ:/ (ú is /u:/). Listen to the very first line of their national anthem, Lofsöngur (Song of praise). It begins Ó, guð vors lands! Ó, lands vors guð! (O, God our land! O, land our God’s!):

    I hear neither a ’v’ nor a diphtong. It’s otherwise true that the pronunciation of modern Icelandic differs a lot from the medieval days. Lexicon and grammar have been quite stable though and thus they can read their ancestors’ Sagas but would face difficulties talking with Snorri Sturlason.

  14. I can think of many other concepts expressed as tetragrammata in numerous languages. For example, Italian cane, French chien, German Hund, Polish pies, Ancient Greek kúōn, Estonian koer, (Romanised) Arabic kalb, Xhosa inja, Swahii mbwa, Yucatec Maya peek’, etc. Is it so because every dog has four paws? Of course there are exceptions like English dog, but we have already established that English is strange.

  15. Self-correction: chien has five letters, so let’s use Hausa kare instead. I suppose languages with five-letter ‘dog’ words (also Spanish perro, Hungarian kutya, Finnish koira, Turkish köpek . . .) count all major appendages rather than just limbs.

  16. Piotr: As Father Arator pointed it out back in 1590, from the point of view of tetragrammaton-counting, Hungarian has the strange peculiarity that two letters may be counted as one. ʻTyʻ, which is in fact one sound, was written in the pre-Christian runaic script with one sign, so ʻkutya’ is no exception from the rule.

  17. Studiolum: Yes, I thought of that. I also think chien may be a valid tetragrammaton after all. It’s three phonemes and five letters, which yields 4 as the arithmetic mean.

  18. Stefan, I first read of that -v- in Einarsson’s grammar. I found it odd, so I asked an Icelander to pronounce it for me, which he did. The [v] was very clear and unambiguous, and he was quite aware of it.

  19. ” the Persians Σύρη”

    What is this “Persian” language? In Farsi, Dari and Tajiki, ‘god’ is khoda (of course that would be four letters in the Tajik Cyrillic alphabet that was still centuries in the future). Is this a ghost word, or a Kurdish, Balochi, or whatever form?

  20. Good question, but at this remove it may be impossible to know what he meant by “Persians” (since Persian civilization extended over a huge part of the Middle East and Central Asia — “Persians” could mean in practice “anyone who celebrates Nawruz”) and what he meant by Σύρη (presumably the consonants count for little and the vowels for nothing at all).

  21. Hoffman’s Lexicon universale reports the same about various tetragrammata, in a reduced form, as a quotation from Ficino’s Argumentum in Platonis Cratylum, writing the Persian name ʻSyre’ in Latin characters instead of Greek (although he can use Greek characters where needed). I think it would not be useless to check the respective notes in Ficino’s modern critical edition (or also in an old, but well commented edition, as they often quote more ancient historical material than modern ones). In any case, this name seems to be attributed to ancient Persians, not Arator’s or Ficino’s contemporaries.

  22. Here you have Ficino in English (explaining the tetragrammaton in a different way), quoting the Persian ʻSyre’, but without notes: http://tinyurl.com/mv6dugk

  23. Stefan Holm says

    Good enough, Y, attestation from a native is of course irrefutable. A general fronting of /u:/ to /ʉ:/ or /ʏ:/ is common to Norwegian, Icelandic and Swedish and thus unproblematic. But an intermediate /v/ seems phonetically ‘uneconomical’ (a /w/ were easier to understand). I’ll check the next time I bump into an Icelander (which happens to and fro).

  24. And English! /u:/ has been moving forward massively in the last 50 years, though not yet to /y/.

  25. Stefan, try this:


    The /v/ (realised as a labiodental approximant rather than a fricative by this speaker) is very clearly audible. I only know that the pronunciation of guð is historically irregular and exceptional, but I don’t know why.

  26. Stefan Holm says

    English! /u:/ has been moving forward massively in the last 50 years.

    Interesting in deed (I think I have noticed it myself). The front rounded vowels otherwise seem to be quite infrequent around the world (in Africa and the Americas practically absent), as can be seen on this WALS map:

    From some reason neither Icelandic nor Swedish is on the map (or on the following list of 562 languages). In western Euroasia it is basically present in French, Breton, German, Scandinavian, Turkish and Fenno-Ugric dialects.
    Would English be added within the next few decades?

  27. Stefan Holm says

    Thank you Piotr. Your Icelandic namesake Petur, 75 certainly has a dental or labial fricative there. Nice to know that I can always turn to our southern neighbours for information how to pronounce po islandzki.:-)

  28. I think it’s only a central rounded vowel now, but since there is no front-back contrast in English high rounded vowels, any of [u], [ʉ], [y] will be heard as /u/.

  29. John!:

    In my ears the real problem with English speech versus spelling is its use of myriads of diphtongs. They are’nt just there in the Latin alphabet. In Scandinavia we have been able to cope with our peculiarities through the diacritics (å, ä and ö). On the other hand the letters ‘c’, ‘q’, ‘w’, ‘x’ and ‘z’ are totally redundant in Swedish (well, in an anglophone world the ‘w’ may be needed).

    But what would you do about the different spellings of ‘hair,’ and ‘hare’ , ‘great’ and ‘grate’ , ‘knew’ and ‘you’, ‘would’ and ‘wood’? And the use of the same sound value (long or short) in ‘pool’ and ‘pull’, ‘feel’ and ‘fill,’?

    In Swedish our problems are minor They are about the fricatives; The /x/-sound, which has been extremely productive in Swedish (but can’t be pronounced by most native speechers of English) can be spelled in 45(!) different ways (according to our National Encyclopedia). We could also make use of the cyrillic ‘So a simple exchange of ‘x’ meaning /ks/ to /x/ would help our school children a lot – but of course – making it harder for them to get a grasp of classical litterature.

    So I suppose we’ll have to live with ‘inaff’ being spelled ‘enough’ (or your return to our ‘nog’ or the German ‘genug’)-.

  30. But what would you do about the different spellings of ‘hair,’ and ‘hare’ , ‘great’ and ‘grate’ , ‘knew’ and ‘you’, ‘would’ and ‘wood’?

    I assume your comment was ment for the other thread, but whatever. Hair and hare ar the same in most accents, but not aul: distinguishing them helps a few peeple and duzn’t really hurt the rest, given the “wun spelling, wun sound, but not vice versa” rule. The same is true of great/grate, knew/you. But because there is no /l/ in would, Regularized Inglish changes it to wood, and never mind the homonymy.

    ‘pool’ and ‘pull’, ‘feel’ and ‘fill’

    These differ in quality (what is called “length” in informal English phonology), not just length.

    So I suppose we’ll have to live with ‘inaff’ being spelled ‘enough’

    Enuff is the RI spelling.

  31. The Icelandic -v- occurs only with Guð and its compounds (e.g. this lady‘s name). It’s not an honorific; so what is it? Did the Icelanders desire a proper tetragrammaton?

  32. David Marjanović says

    Ó, guð vors lands! Ó, lands vors guð! (O, God our land! O, land our God’s!)

    That doesn’t work. How about: “O God of our land! O our land’s God!” – still awkward, but grammatical, and preserves the word order.

    But because there is no /l/ in would, Regularized Inglish changes it to wood, and never mind the homonymy.

    Would at least has the usual etymological excuse, compare German wollte. But there never was any kind of /l/ in could, compare German konnte; the spelling is purely from analogy to would!

  33. About the “Persian” name Syre, I believe it could be the Avestan sūra, which means “powerful, mighty, strong, brave” (it corresponds to the Sanskrit śūra). It is an epithet, that is applied to Mithra in Avesta and also possibly is present in the name of the Scythian Apollo – Goito-syros (according to Herodotus, IV, 59).

    I have encountered other similar lists of “divine tetragrammata amongst the different people” and somewhere (I don’t remember where exactly) the name “Abgd”, ascribed here to the Saracens, is related to the “Philosophers” (taken as “nation”?). In my opinion this means the Pythagoreans and their famous “Tetraktys”, i.e. the numbers 1-2-3-4, designated with the Greek letters aplha-beta-gamma-delta, hence “Abgd”.

    Also curiously the ancient Turks also wrote the Altaic word for “god” – Tengri – with four runic letters (phonetically: “Tŋri”).

  34. About the “Persian” name Syre, I believe it could be the Avestan sūra, which means “powerful, mighty, strong, brave” (it corresponds to the Sanskrit śūra). It is an epithet, that is applied to Mithra in Avesta and also possibly is present in the name of the Scythian Apollo – Goito-syros (according to Herodotus, IV, 59).

    Thanks, that seems very likely.

  35. Wiktionary says the Quechua for ‘God’ is the obvious borrowing Dyus ([o] in Quechua is an allophone of /u/), which is a tetragrammaton but has nothing to do with “Peruvian” zimi. In Aymara it’s Tatitu. Other tetragrammata on that page are Bafiaxx: Bɛ̀ll, Baluchi ہدا ‎(huda) or خدا ‎(xuda), Breton: Doue, Hawaiian: Akua, Hawaiian: Akua, Istriot: Deîo, Kabardian: Тхьэ ‎(Tḥă), Kamba: Ngai, Kikuyu: Ngai, Latgalian: Dīvs, Luhya: wele, Maltese: Alla, Nauruan: Gott, Ngazidja Comorian: Mngu, Portuguese: Deus, Somali: Eebe, Yaqui: Lios.

  36. David Marjanović says

    Now we have discovered the true meaning of the Russian orthographic reform of 1918. It was the Bolshevik design to change Богъ to Бог and ruin the Trinity.

    The reform theoretically belongs to the “bourgeois” revolution of a few months earlier.

  37. David L. Gold says

    “Now we have discovered the true meaning of the Russian orthographic reform of 1918. It was the Bolshevik design to change Богъ to Бог and ruin the Trinity.”

    “The reform theoretically belongs to the “bourgeois” revolution of a few months earlier.”

    Actually, even earlier:

    “Under the leadership of scholars F.F. Fortunatov and A.A. Shakhmatov, the movement towards reform [of Russian spelling] gradually took shape. On April 12, 1904, a Commission on the Question of Russian Orthography, organized by the Academy of Sciences, met to discuss the matter, voting in favor of reform and the abolition of some of the letters mentioned above. It then set up a subcommission to work out further details. With a long interruption following the 1905 revolution, the subcommission finally published its Resolutions of the Orthographic Subcommission in 1912, which was to become the basis for the actual implementation of reform in 1917” (Eugenia Sokolskaya, “Seize the Ъ: Linguistic and Social Change in Russian Orthographic Reform,”

    There was Russian Orthodox opposition to abolition of letters of the Church Slavonic alphabet (see that article) but there seems to be no evidence that it was motivated by a fear that “Богъ would be changed to Бог and ruin the Trinity.” If there was such a fear, I have not found the evidence.

  38. Good historical summary, but you’re responding seriously to what was a joke.

  39. David Marjanović says

    Actually, even earlier:

    Yes, but “the actual implementation of reform in 1917” was what I meant.

Speak Your Mind