In the course of researching the etymology of the word "Australia", I was trying to find the Latin words for north and south (the cardinal directions). I found some websites that translate north as "Septentrionalis", but I understand this to refer to the seven oxen, or what we today call the Big Dipper, as it is in the northern sky. Other websites translate north as "boreas" or "aquilon" though I think that the former is actually a Greek wind god and the later is his Roman name.

Now of course in English the words north, south, east, and west have no transparent etymologies; they are just the words for the directions. But it seems that Latin and Greek had direction words that were derived from other things (gods, stars, oxen, winds).

This line of thought leads to a question: historically and cross-linguistically, how did terms for the cardinal directions arise? Or differently stated, is it the case that the cardinal direction terms are derived from similar processes across different language families?

  • 1
    The only Latin words for north, east, south, and west that we know are all metonymical: winds and and the movement of the sun (e.g. oriens = "up-coming" = east). No basic words as we have them are known. Auster is the (hot) south wind, metonymically the south, and that is what Australia is named after.
    – Cerberus
    Commented Nov 27, 2011 at 1:11
  • 1
    North is septentrio or boreas, south is meridies or auster. (Which to use at what times is a question all to itself!) Yes, the words do have other original meanings - not everything can be a basic root word.
    – Muke Tever
    Commented Nov 27, 2011 at 15:27
  • 2
    I believe that a good answer to this question is possible. It's an interesting particular case of language change and how new words are formed. Commented Apr 27, 2012 at 13:37
  • 1
    @MukeTever I think your comment might be the beginning of a good answer to this question. Commented Apr 27, 2012 at 20:21
  • 2
    I see no reason to close this question, it has provoked some very good answers. Commented Dec 12, 2016 at 13:23

4 Answers 4


Direction words arise from the need to coordinate direction. Thus, their nature and usage can vary widely from one language to another. To start, here are a few aspects of a people and their land that can influence the form that direction words / direction morphemes end up taking:

  1. The types of activities requiring coordination (migration, herding, hunting, fishing, gathering, farming, trade relations, war, etc.)

  2. The geographical features of the area (coastline, a central river, tributaries, woods, islands, steppe, mountains).

  3. The current state of inherited cultural knowledge about direction (including mythology; this may largely go back to (1) and (2) from prior states of the language).

The systems that thus arise may be influenced to be similar to the common cardinal directions in English, or instead make the primary distinction "upstream" vs. "downstream," "inland" vs. "toward the coast," or "direction towards" vs. "away from a certain very distant landmark." Taken beyond that landmark, a person might not be as effective at communicating navigation. (Although, technically, the same thing can happen in English: explain which way to go from the North Pole!)

In Orientation Systems of the North Pacific Rim by Michael Fortescue, a close examination of the orientation systems of Wakashan, Tsimshianic, Haida, Tlingit, Eyak, Athabaskan, Eskimo-Aleut, Tungusic, Nivkh, and Ainu leads to the enticing conclusion that by conceiving of a language's orientation system diachronically, looking at etymologies of direction words and inspecting any inherently paradoxical methods of expressing direction, we can evidence a hypothesis as to whether a language is relatively new to a region (say, having arried within the past 2000 years), as well as perhaps find out where they came from!

For example, if people in a dialect continuum settle a turning peninsula, the meaning of a direction word at the source may be different from at the tip. Similarly, whereas most IE languages use the PIE root *aus- "to shine, dawn" for "east," the Latin word australis "south" may be from the same root. This is "perhaps is based on a false assumption about the orientation of the Italian peninsula, 'with shift through "southeast" explained by the diagonal position of the axis of Italy'...Or perhaps the connection is more ancient, and [directly] from PIE root *aus- 'to shine,' source of aurora, which also produces words for 'burning,' with reference to the 'hot' south wind that blows into Italy." OE

To address your question about English (all quotations from the Online Etymology Dictionary, OE):

  • north < norð < *nurtha possibly derives from PIE *ner- "left, below" "as north is to the left when one faces the rising sun (cf. Skt. narakah 'hell,' Gk. enerthen 'from beneath,' Oscan-Umbrian nertrak 'left')." OE

  • south < suð < *sunthaz is "perhaps related to base of *sunnon 'sun,' with sense of "the region of the sun." OE

  • east < east < *aus-to-, *austra-, "from PIE *aus- 'to shine,' especially 'dawn' (cf. Skt. ushas 'dawn,' Gk. aurion 'morning,' O.Ir. usah, Lith. auszra 'dawn,' L. aurora 'dawn,' auster 'south'), lit. 'to shine.'" OE

  • west < west < *wes-t- "from PIE *wes- (source of Gk. hesperos, L. vesper 'evening, west'), perhaps an enlarged form of root *we- 'to go down' (cf. Skt. avah 'downward'), and thus lit. 'direction in which the sun sets.'" OE

Some languages use non-compound words for the ordinal (secondary) directions. For example, Finnish (with help from Wiktionary and Finnish Wikipedia):

  • luode "northwest," possibly the same etymology as identical luode meaning "ebb / low tide," loan from a Germanic language, cognate of German Flut, Swedish flod.

  • pohjoinen "north," from pohja "bottom," is because the sun is in the north when it's underneath the horizon, possibly also because the back of a dwelling should be facing the north so as to maximize warmth.

  • koillinen "northeast," from koi "dawn."

  • itä "east," possibly related to itää "to germinate," that the sun grows in the east.

  • kaakko "southeast," equated with kaakkuri "red-throated diver (loon)" and kuikka "black-throated diver (loon)." Compare Latin ornithias "bird-wind," the spring wind that brings the birds.

  • etelä "south," antonymously to pohjois, is the direction in which the front of the house should face. Compare eteen "to the front." The Estonian cognate edel means "southwest."

  • lounas "southwest, lunch" indicates the direction the sun is in at lunchtime. The Estonian cognate lõuna means "south."

  • länsi "west" may have to do with the day being "flattened" (läntätään ~ litistetään) as the evening arrives.

  • 4
    +1 Nice mention of Finnish! :D I like elaborated answers. It shows interest.
    – Alenanno
    Commented Apr 29, 2012 at 10:55
  • 2
    @MarkBeadles We have changed that. Now it's possible to ask about single languages. :P But it's still more interesting to make comparisons ahah
    – Alenanno
    Commented Apr 30, 2012 at 10:35
  • 2
    @Alennano The change is probably a good one, since we are hurting for questions. Commented Apr 30, 2012 at 11:25
  • 1
    @MarkBeadles Yes. Our visits are slowly increasing, but we can do more. By the way, any possible change you might think of the FAQ, feel free to propose it on Meta. :D
    – Alenanno
    Commented Apr 30, 2012 at 12:39
  • 1
    Thank you Daniel. Your terrific and detailed answer is very informative, and as it best answers the question as it is currently worded, I will mark it as the accepted answer. @Alenanno: Might I be able to reask the original question now that the scope of the site has changed?
    – dotancohen
    Commented Apr 30, 2012 at 17:57

I will look specifically at the western Classical origins of these terms. The Ancient Greeks in particular had extensive mythology and naming associated with all the Winds (Ἄνεμοι Anemoi) and directions, not just the cardinals. The Greeks were a seafaring people and wind direction was central to their lives.


L septentriō (adj. septentriōnalis) = septem "seven" + triō, that is "the seven plough-oxen (stars of Ursa Major). triō (pl. triōnes) is problematic. Most sources give this as meaning "plough-oxen", but this term for plough-oxen is used nowhere else in Latin. A few sources posit that triō < PIE *(s)tē̆r- "star" with loss of initial s as in Indic, and that later mythology led people to reinterpret the root as "oxen". The transparent meaning of this word to the Romans was "in the direction of the constellation the Plough (Ursa Major)".

L boreās (adj. boreālis) was also used to mean "north" or "North Wind" and was a direct borrowing from Greek Βορέας. The Romans also called this Wind Aquilō; this latter word is of unsure etymology. There have been attempts to relate it to aquila "eagle", aquilus "dark" and aqua "water", viz. "rainy wind".

Gk Βορέας boreas "north, the North Wind" was a Greek word also of unsure etymology. It has cognates in other Balkan and Slavic languages such as Alb borë "snow", Srb бура "cold north wind". It is often said to come < PIE *gʷor- "mountain". This very likely is a reference to a North wind, cold and perhaps arising from mountains, that was prominent in these people's original homeland.


L merīdiēs (adj. merīdiōnālis) meant "noon, midday" < medius “middle” + diēs “day”. Since the sun is in the South at midday in the Northern hemisphere, this word is self-explanatory.

L auster (adj. austrālis) was the Latin name of the South Wind and the South. Now here is a fascinating bit of history and its relation to language change. Most scholars believe that auster < PIE *-aus "shine" - which is the same root that gives rise to the words for "dawn" and "east" in other IE languages! How can the same root be used for different cardinal directions in sister languages? One theory is that since the Italian peninsula runs diagonally NW-SE, the word for "east" shifted to mean "south" since both were in the direction of the Tyrrhenian Sea. Another theory is that since the lands to the South were burning hot, or alternatively since the sirocco was hot, the root *aus- referred to this heat.

Gk νότος notos was the South wind of the Greeks. I have no idea what its etymology is.


L oriēns (adj. orientālis) was the usual Latin term. The meaning was transparent in Latin: "rising", viz. "in the direction of the rising sun". oriēns is the present participle of the deponent orior, "rise" < PIE *or- O-grade of "move"

Gk ἠώς eos was used in Greek and also meant "dawn", which is cognate with Latin aurora and Germanic east. ἠώς < PIE *h₂ewsṓs/*h₂ausōs < *aus- "shine". This is conjectured to be a reference to the shining dawn; but see 'south'.


L occidēns (adj. occidentālis) was the usual Latin term. The meaning was similarly transparent in Latin: "going down/setting", viz. "in the direction of the setting sun". occidēns is the present participle of occidō, "fall/go down" < ob “towards/facing” + cadō "fall” < Proto-Indo-European *ḱad- “fall”.

L vesper "evening" was also used to mean "west" in reference to the setting sun.

Gk ἕσπερος hesperos was found in Greek, cognate to Latin vesper and Germanic west. ἕσπερος < PIE *wesperos/wekeros "evening" < *wes "wind, blow" + *pero "source". In origin this may have been something like "the direction from which the wind blows"; one can imagine that this is possibly a reference to the prevailing winds in the PIE urheimat.

  • 1
    Thank you Mark, there is some very interesting etymology in there! I pronounce "ε" as "e" in then, is this incorrect? I see that you transliterate it as "he".
    – dotancohen
    Commented Apr 30, 2012 at 18:00
  • 1
    @dotancohen The epsilon indeed is transliterated as e -- however when it has a "rough breathing" diacritic as in , that is usually transliterated with a preceding h in English. Commented Apr 30, 2012 at 19:22
  • Thank you Mark. I figured that the diacritic marks were the key, but I don't know where to research their meanings.
    – dotancohen
    Commented May 1, 2012 at 7:50
  • 2
    Damn, I just binged the internet over the past hours. It started after discovering the word "Thrall", somehow landing on "Werewolf", then looking for "Wer", then "War", "Werra", "Guerra", "Gere", then proposing my brain "What abour Ger-mans? Guerra-Men? Men of War?". As a sidekick I wondered about latin "australis" (southern) and "austr vegr" (east way). I then awoke in some PIE articles and ultimately landed here. On the way through, I've learned that Polish/Slavic "niemiecki" for German basically means "dumb" (what to expect after en-slave-ing them?). Linguistics are some good Rock'n'Roll. Commented Nov 17, 2017 at 8:34
  • 2
    @SebastianMach - +1, but "Niemiecki" is not dumb, but "mute" - not able to speak (speaker's language) in all Slavic languages (I know of). Commented Jan 8, 2021 at 23:34

To add to the above, in PIE

  • South: *aug-
  • North: *(s)kewer- (also North wind)
  • East: *aus(t)-

and the word for the West is unknown.

  • 7
    This kind of discussion is always more helpful with references. Commented Apr 30, 2012 at 11:27
  • 1
    Can you elaborate this answer a bit? Commented Dec 12, 2016 at 13:21
  • @jknappen what do u mean?
    – Anixx
    Commented Dec 12, 2016 at 14:08
  • 1
    @Anixx: What sources on PIE did you use, what are later witnesses of the PIE roots. BTW, I also wonder about the discrepancy between *aus (stated by OED) and *aug in your answer. Commented Dec 12, 2016 at 14:15
  • 1
    @jknappen my answer is very old, I do not remember the sources. In general, the root a̯u̯es- meant (golden) glow. Thus we have a̯u̯esom gold, a̯u̯esos dawn, a̯u̯estrom east, a̯u̯ostus "abode", place to pass the night till morning. Country names of Austria and Australia came from this root, as well as English "East" as well as "aura". Regarding the root for south, I cannot find a source now, but it possibly gave Slavic "yug" as in Yugoslavia. There is a hypothesis also that they could call south with the word for right hand side descnos.
    – Anixx
    Commented Dec 12, 2016 at 17:21

By giving the Sanskrit Equivalent, I wanted to show you the Etymology of Nordri, Sudri, Austri, Vestri- with Sanskrit Shudra,kShatri, Veshya .

My point is Sanskrit stands to be the oldest Indo-European language . PIE is not a language rather a hypothesis.PIE was never a language and not even one word belonging to PIE has been found. Isn't it time that the objective philologist start paying attention to the antiquity of Sanskrit,

  • 1
    ... because the past two centuries of Indo-European linguistics has somehow ignored Vedic? (Not Sanskrit, that's later.) Commented Jul 19, 2018 at 1:46
  • What does it mean for a given language to be "old"? Or the "oldest"? Commented Jul 19, 2018 at 8:03

Not the answer you're looking for? Browse other questions tagged or ask your own question.