The colors of the seas and the directions

« previous post | next post »

Tim Leonard just came across this article from five years ago that hypothesizes a common origin for the names of the Red Sea, Black Sea, and (in Turkish) White Sea and Blue Sea:

"It Works for the Turks: A Colour for Each Direction"
Frank Jacobs, Big Think (3/6/16)

Peter Golden has often talked to me about how the directions were associated with certain colors among the Turks, and how these associations had powerful symbolic and intellectual functions and revealed much about the history of the Turks and their interactions with other peoples.  By looking at some of the ideas presented in this article, I'm hoping that we can elicit further details from Peter and colleagues who are familiar with color associations in other cultures.

What colour is the sea? Why, blue of course (1). Unless you're Homer. In which case your numerous references to the Aegean as the "wine-dark sea" will puzzle scholars for centuries to come (2).

There are other curious exceptions. If you're Turkish, the question What colour is the Mediterranean? is its own answer. The Turks call that body of water Akdeniz, which literally means: the White Sea.

But then of course English also has its coloured seas – four, to be precise, although none of the toponyms is native to the English language. One also is a White Sea, though not quite blessed with the same balmy climate as its Turkish namesake. A direct translation from the original Russian (3), the white in this sea's name refers to the ice floes that block shipping to and from Arkhangelsk, the region's major port, for at least half the year.

The Yellow Sea is another direct translation. Called Huánghǎi (黄海) in Chinese, the sea gets its name from the sand particles that travel down the Yellow River, or Huánghé (黄河), from the Gobi Desert, and can turn the sea's surface golden yellow.

The Black Sea is rich in iron sulfide, allowing very little else but sulfur bacteria to thrive. Hence the darkness of the sea's sediment, and of its waters when stirred. The oldest recorded name for the sea, dating from around 500 BC, already refers to its colour: the Achaemenids called it Axšaina, (Persian for 'black').

Greek popular etymology transformed the Persian word into axeinos ('inhospitable') when Greek settlers first arrived on the sea's shores, and into euxinos ('hospitable') when they became more numerous in and familiar with the region.

The last coloured body of water is the Red Sea, separating Africa from Arabia. The Greeks already called it Red Sea, or Erythra Thalassa (Ερυθρὰ Θάλασσα) (4). One theory is that the red of the name refers to the seasonal blooming near the water's surface of a red-coloured cyanobacterium called Trichodesmium erytrhaeum – sometimes also known as sea sawdust.

But there is another theory. Certain Asiatic languages use colours to refer to cardinal directions. In his Histories, Herodotus on one occasion refers to the “Erythrean or southern” sea, using the terms interchangeably. Could it be that the Red Sea owes its name to something else than bacteria – and perhaps the Black Sea too?

Let's go back to the curious Turkish name for the Mediterranean. Turkish is one of several languages and cultures to associate the cardinal directions with colours. White, you guessed it, is linked to west, as the Mediterranean lies to the west of Asia Minor, homeland of the Turks. Two other names also fall into place: north is black, south is red. And indeed: the Black Sea – Karadeniz – is to the north of Turkey, the Red Sea – Kizildeniz – to the South.

What about east? In Turkish, that direction is associated with blue. But there is no Gökdeniz ('Blue Sea') to the east of Turkey. The likeliest candidate is the Caspian Sea, landlocked between Russia in the north and Iran in the south, flanked by the Caucasus region in the west and the 'stans in the east. But it is called Hazar denizi in modern Turkish. Literally translated, the 'Khazar Sea', after the disappeared Turkic tribe that famously converted to Judaism around the 9th century.

However, on Emanuel Bowen's 1747 Map of Iran, a northern inlet of the Caspian Sea is labelled the Blue Sea. The area has been known afterward as Tsesarevich Bay and, in communist times, as Komsomolets Bay. It is currently called Dead Kultuk. Could its former name be a residue of an older Turkish toponym, one that covered the entire sea? Then again, other sources seem to suggest that the Aral Sea may have been the Turks' orlginal Gökdeniz.

The question of coloured cardinal directions is not just of interest from a historical or linguistic point of view. Linking Turkish culture to ancient toponyms provides a proximate location for Turkish ancestral lands. In a region riddled with ancient grudges about territorial loss and conquest, such archaeo-linguistic evidence of residential antiquity is highly valuable.

(1) From Blackadder, recall Baldrick's pathetic attempt at lexicography by defining the letter C as: "Big blue wobbly thing that mermaids live in". 

(2) It has been suggested that Homer was colour-blind, that ancient Greek lacked a word for 'blue', that marine algae gave the Aegean that particular hue, that it referred to dust-induced red sunsets, and even that Greek wine in antiquity was… blue! Indeed, in a 1983 letter to the science journal Nature, two Canadian scientists proposed that the water used by the Greeks to dilute their wine contained alkalines of such quality and in such quantity that it turned the originally red-coloured drink blue. Likeliest still is the standard explanation: that of poetic licence.

(3) Byeloye More (Белое море). Curiously, the related Slavic languages of Serbian and Bulgarian use their equivalent to describe… the Aegean.

(4) Also the origin of the name of the modern state of Eritrea, thus, literally: "Redland".  [VHM:  A text that I have studied for many years is the Periplus of the Erythraean Sea, a logbook for sailing itineraries and ethnological and commercial observations concerning the ports of the Erythraean Sea (Greek: Ἐρυθρὰ Θάλασσα, Erythrà Thálassa, lit., "the Red Sea"), the Gulf of Aden between Arabia Felix and the Horn of Africa, with extensions eastward as described in the following quoted paragraph.]

The Periplus of the Erythraean Sea (Ancient Greek: Περίπλους τῆς Ἐρυθρᾶς Θαλάσσης, Períplous tē̂s Erythrâs Thalássēs, modern Greek Períplous tis Erythrás Thalássis), also known by its Latin name as the Periplus Maris Erythraei, is a Greco-Roman periplus written in Koine Greek that describes navigation and trading opportunities from Roman Egyptian ports like Berenice Troglodytica along the coast of the Red Sea, and others along Horn of Africa, the Persian Gulf, Arabian Sea and the Indian Ocean, including the modern-day Sindh region of Pakistan and southwestern regions of India.


The traditional Chinese color/direction scheme seems to match the Turkish one, with blue-green to the east, red to the south, white to the west, and black to the north.  Chinese adds yellow in the center, and identifies the five colors and directions with the five metaphysical phases / elements.


Selected readings


  1. ktschwarz said,

    April 28, 2021 @ 4:48 pm

    A more scholarly reference for the color/direction symbolism, with further references: Black Sea by Rüdiger Schmitt, at Encyclopædia Iranica.

    Also see: (this book is on
    F. de Blois, "The name of the Black Sea." In: Maria Macuch, Mauro Maggi, Werner Sundermann (eds.), Iranian languages and texts from Iran and Turan. Ronald E. Emmerick memorial volume. Wiesbaden: Harrassowitz (2007).

  2. Martin said,

    April 28, 2021 @ 5:21 pm

    "the disappeared Turkic tribe that famously converted to Judaism around the 9th century."

    Shaul Stampfer questions the veracity of the story that the Khqzarian royal family and the Khazarian nobility converted to Judaism in his article “Did the Khazars Convert to Judaism?” (Jewish Social Studies, vol. 19, no. 3, Spring-Summer 2013, pp. 1-72).

    The full article (73 pages) may be read free of charge here:

    1. the publisher's website:

    2. Research Gate (downloadable)

    3. Jstor (downloadable)

  3. Hiroshi Kumamoto said,

    April 28, 2021 @ 6:21 pm

    ἐπὶ οἴνοπα πόντον

    On note (2) above, see:

    "Why on Earth would my red wine turn blue when I ran water into the glass?" ( )

    "Why does red wine turn blue when I wash out a wine glass with water?"
    ( )

    and, of course, the Wikipedia article:

    "Wine-dark sea (Homer)" ( ).

  4. Peter B. Golden said,

    April 28, 2021 @ 6:49 pm

    Беларусь/Belarus' (White Rus') is the western zone of the Eastern Slavs.
    In Muscovite documents translated from Turkic, the Muscovite ruler was often term Белый царь (the "White king,") i.e. the "Western Ruler"), a geographical reference. Southwestern Ukraine (roughly Halychyna, Polish Galicja) in modern Ukraine was termed Ukr. Червона Русь, Russ.Червонная Русь, Polish Czerwona Ruś. Červonnyj etc. denotes "red."
    During the Russian civil war, in addition to the Конармия made famous in Babel's "Red Cavalry" stories, there was a very large force termed Червонное казачество (Red Cossacks) that fought across the south (mainly Ukraine). My paternal grandmother's younger brother served in it – and told me many stories, when I visited with him in Leningrad in 1968.

  5. Victor Mair said,

    April 29, 2021 @ 10:07 am

    From Michael Witzel:

    I have dealt with the color scheme for the directions of the sky in my first paper, almost 50 years ago:

    "jav. apāxədra im System der avestischen Himmelsrcihtungsbezeichungen.
    Münchener Studien zur Sprachwissenschaft, 1972, p.163-191."

    see footnote 19:
    “….Reste eines weiteren Bezeichnungsystems für die Himmelsrichtungen, das wie das chinesche mit Farben arbeitet, (O. = grün/hellblau, S. = rot, W. = weiss, N. = schwarz) liegen wohl wenigstens im Iranischen vor, worauf schon H. Hirt, Schwarzes und Rotes Meer, Geogr. Zeitschr. XXXII, p.430 hingewiesen hat. Dem ist noch der Name 'weisses Indien' für die unter den Seleukiden an das indische Mayuryarreich abgetretenen Provinzen Arachosien usw. hinzuzufügen (vgl. Marquart, Untersuchungen zur Geschichte Irans, I p. 31), die damit die westlichsten Provinzen Indiens bildeten. Auch der Aralsee soll früher den Namen 'blaues Meer’ geführt haben (s. Tolstov, Auf den Spuren der altchoresmischen Kultur p.57).


    ” … remnants of another designation system for the directions of the sky, that works, just like the Chinese one, with colors (East = green/light blue, South = red, West = white, North = black), at least appear in Iranian, which has already been pointed out by Hirt, “Black and Read Sea", Geogr. Zeitschrift XXXII, p. 430. To this has to be added the designation ‘White India’ of the provinces Arachosia etc , ceded by the Seleucids to the Indian Maurya realm, (cf. Marquart, Untersuchungen zur Geschichte Irans, I p. 31), which thus were the westernmost provinces of India. The name of the Aral Sea is said to have been ‘blue sea’ (see Tolstov, Auf den Spuren der altchoresmischen Kultur p. 57).


    Add Bela-rus, ‘white Rus’, formerly called White Russia in English etc. = ‘western Russia."

    And I recall that the Aztecs have a similar color scheme….

  6. Jerry Friedman said,

    April 29, 2021 @ 1:19 pm

    The Navajos' correspondence is white—east, blue (turquoise)—south, yellow—west, black—north. The directions also correspond to mountains, deities, and ornamental substances.

    Isleta Pueblo's is white—east, red—south, blue—west, yellow—north. Black and speckled are up and down, but the Facebook post I linked to doesn't say which is which. Those are the six colors of corn (maize). I think many other Pueblo tribes have similar correspondences, often including mountains.

  7. Daniel Barkalow said,

    April 29, 2021 @ 1:40 pm

    There's a lot of really obvious evidence that what portions of the spectrum are assigned color names of various degrees of specificity is a matter of culture rather than physiology or physics. There just isn't a correct answer to the question of how many colors there are in the rainbow between red and green or between green and violet, let alone where the transitions are. Even among English-speaking Americans, people who do graphics and the general public disagree in opposite directions on whether orange is a separate color and on whether cyan is a separate color (and we teach children that indigo is a separate color from blue, but people who distinguish those colors call them blue and cyan, respectively). It also seems clear that people's judgements of "a different shade of the same color" is based on the language they use rather than some non-linguistic category boundaries.

  8. Andreas Johansson said,

    April 29, 2021 @ 3:57 pm

    From an Anatolian perspective, the Mediterranean is surely at least as southern as western. The name might make more sense if coined before Turkic-speakers had penetrated that far west.

    I've long thought it funny, BTW, that Turks and Russians agree that the sea between them is the black one, and the one on the opposite side of their own country is the white one.

  9. R. Fenwick said,

    April 29, 2021 @ 9:25 pm

    A well-known association of colours with cardinal directions is also seen in Classic Maya (lak’in "east" = chak "red", xaman "north" = sak "white", chik’in "west" = ik’ "black", nohol "south" = k’an "yellow"), forming part of a much larger quadripartite cosmology that also incorporates deities, astronomical phenomena, animal sacrifices, and the Year Bearer day names:

    East: k’in "the Sun", chij "deer", god Hobnil, day Ik’
    North: uh "the Moon", kutz "turkey", god Kan Tziknal, day Manik’
    West: ak’ab "the night", huj "iguana", god Sak Kimi, day Eb’
    South: chakek’ "Venus", chay "fish", god Hosan Ek’, day Kab’an

    For the Maya, their fifth basic colour term, yax "grue", also represents the centre of the four directions, probably representing the earthly, living, growing world contained within the celestial sphere. I suspect the four colours come from the astronomical phenomena associated with them: red for the Sun (viz. sunrise), white for the Moon, yellow for Venus, and black for the night sky. Black is also the colour associated with Xibalba, the Maya underworld, said to be located in the far west.

    (Intriguingly, the three colours the Maya system shares with the Navajo system are rotated anticlockwise ninety degrees: where Navajo white, black, and yellow correspond to east, north, and west respectively, in Classic Maya these colours represent north, west, and south. Whether this is more than chance I don't know.)

  10. R. Fenwick said,

    April 29, 2021 @ 10:26 pm

    @Daniel Barkalow: There's a lot of really obvious evidence that what portions of the spectrum are assigned color names of various degrees of specificity is a matter of culture rather than physiology or physics.

    "Rather than"? Do you have a citation for that "really obvious" evidence? Certainly once one reaches beyond a certain point the proliferation of colour names begins to fall into straightforward derivation from objects that are indeed culturally bound – "azure", "teal", "violet", "mustard", "fuchsia"; even "orange" and "purple" were originally so, and the development of complex colour systems in languages that provably used not to have them is very often the result of cultural diffusion too. But when you're talking about the development of the basic systems, the ones that most speakers of a given language agree are essentially irreducible to shades of others, there's ample evidence that physiology is indeed involved at least to a certain degree. As such, the development of colour terms is a result of both physiology and culture.

    For instance, the few minimally basic systems that exist are all divided along an axis roughly corresponding to light and dark, or perceived brightness, which is to say, the range between red or orange and yellow-green: for instance, Bassa (Sierra Leone) ziza "white/yellow/orange/red, light, warm" vs. hui "black/blue/green/purple, dark, cool"; Jale (Papua New Guinea) hóló "white/yellow/orange/light green" vs. siŋ "black/blue/dark green/purple/red". This happens to coincide with the zone of maximum activation of retinal photoreceptor cells, due to the large overlap between the optical sensitivity curves of the red and green cone cell rhodopsins. And the next simplest systems generally begin making their higher-level divisions within the same region: red/white/black (though in fairness, the primacy of "red" is also culturally influenced by the importance of blood), then red/yellow/white/black or red/green/white/black, then red/green/yellow/white/black. That green and blue are commonly not distinguished suggests that the lack of brightness makes the distinction less salient. Even more often, words for purple are mostly etymologically transparent references to real-world objects (English violet, Latin purpura, Turkish mor, Hungarian lila, Japanese murasaki, Malagasy volomparasy, etc.), strongly suggests that these are late developments that came into common use only once dye technology became sufficiently complex to require distinguishing one shade of "blue" from another. "Purple" is almost never separated from blue or red without some clear and relatively recent cultural influence, suggesting that the blue/purple distinction is far less physiologically salient than the green/yellow.

  11. Anthea fleming said,

    April 30, 2021 @ 6:12 am

    I recollect reading comment from an amateur landscape painter who had been painting in Sicily. His village neighbours agreed that he could draw, but objected to his use of blue pigments for the sea. In local songs and conversation it was always black. Votive primitive paintings of ships, boats,
    shipwrecks and marine escapes in local churches always showed a black sea.

  12. Gerald Friedman said,

    April 30, 2021 @ 8:28 am

    Daniel Barkalow: Even among English-speaking Americans, people who do graphics and the general public disagree in opposite directions on whether orange is a separate color and on whether cyan is a separate color (and we teach children that indigo is a separate color from blue, but people who distinguish those colors call them blue and cyan, respectively). It also seems clear that people's judgements of "a different shade of the same color" is based on the language they use rather than some non-linguistic category boundaries.

    For another example within a culture, some of my co-workers have gotten emergency-response training and have bright-colored vests to put on in an emergency. We've been told to look for green vests, which surprised me, because I would have called them yellow. I thought that might be because of my deuteranomaly, so I've asked a few people—there's not much to do during fire drills—and some said "green" and some said "yellow". I think somebody said "yellow-green".

    One could also look at descriptions of flowers in catalogs, birds with "purple" in their names, etc.

  13. Alexander Browne said,

    April 30, 2021 @ 11:01 am

    @Gerald Friedman: When the "yellow vest" protesters were active in France, I always thought they should be called green myself. (Maybe "neon green"?) My partner disagreed, and said they're yellow, which didn't surprise me since we often disagree whether darker, mustard-y shades are yellow or green.

  14. TC said,

    April 30, 2021 @ 11:06 am

    Isn't it virtually certain that the system originates from the Chinese? The chinese directional-color symbolism goes back to the Iron Age to the school of Wuxing–philosophers interested in astrology and alchemy.

  15. Victor Mair said,

    April 30, 2021 @ 6:37 pm

    More from Michael Witzel:

    I have checked on the Aztecs. In my myth book (2012) “The origins of the world’s mythologies" we find:

    p.498, Note 479:

    With the Aztec, east = white, south = blue, west = red, north = black (Lehmann 1953 : 42 sq.). The Hopi and Navajo, however, have red/copper/east, blue/silver/south, yellow/gold/ west, and yellowish-white/mixed mineral sikyápala /north (see §2, n. 262). The Navajo also have black as the fourth color (north); see Locke 2004 : 58; §2, n. 262.

    On the other hand, the Chinese and Old Iranian colors for the directions of the sky are different:

    east = blue/green, south = red, west = white, and north = black (Witzel 1972).

    The traditional colors for the three or four classes of the Indo-Europeans are white for the priesthood, red for the nobility, blue/green for the “people,” and sometimes, a fourth class (black) is added; for their interpretation, see Lyle 1990 : 41–47.

  16. R. Fenwick said,

    April 30, 2021 @ 9:59 pm

    @Gerald Friedman: We've been told to look for green vests, which surprised me, because I would have called them yellow.

    If it's the fluorescent lime-green-yellow of what most here in Australia refer to as "hi-vis" (IPA: [ˌhɑe.ˈvɪz]), I'd have called it yellow too (primarily because hi-vis workwear can also be purchased in a deeper and more unequivocal green: cp. e.g. this green), though I suppose I can understand why some people look at it and categorise it as a shade of green as distinct from pink or orange.

    It also reminds me of the old debate: what colour are tennis balls?

  17. R. Fenwick said,

    April 30, 2021 @ 10:02 pm

    @Victor Mair apud Michael Witzel: given the pervasive shared cultural aspects of various Mesoamerican societies in other fields, it surprises me greatly that there's such a discrepancy between the Aztec system and that of the Classic Maya. And given that both Nahuatl and Hopi are (distantly) related within Uto-Aztecan, the fact that the Aztec system is so starkly different from the Hopi system, sharing only blue for south, is also remarkable. I'd hazard a guess that the correlation of red/white/yellow to east/north/west for both the Hopi and the Maya is probably meaningful, as well, though whether independent developments of similar associations or the result of an anciently-diffused Ür-culture is obviously something for researchers better versed in the field than me.

RSS feed for comments on this post