This page contains several maps and information about pacific geography. Several of the maps are unpublished and under construction. They have been made by ANU CartoGIS after my (Hedvigs) requests. ANU CartoGIS is a part of the College of Asia-Pacific at the Australian National University. Their maps are available for free online, here. These maps are meant for publication in my PhD thesis, but will also be available for free to the general public. If you are interested in pacific maps, you are welcome to join the Facebook-group "Maps of the Pacific".

If you find any errors in these maps or have any other comments, you are most welcome to submit them here. Please note that the versions displayed here are already subject to some adjustments based on comments. Previous comments are listed here too, at the bottom of the page. Please note that maps displayed here that were made by other people are not up for revision.

The maps will be finalised under 2019. If you want to use any of the maps on this page, please contact me first and cite appropriately. Maps of ANU CartoGIS maps are provided for use under license CCBYSA - Creative Commons Attribution-ShareAlike 3.0 Australia. If you use one of the maps in a report or publication, please acknowledge CartoGIS Services as follows: "Map provided by CartoGIS Services, ANU College of Asia and the Pacific, The Australian National University".

Problems with previous Pacific maps

Before we start, it's worth describing the problem with the current state of maps of the Pacific that are found in media, wikipedia, text books and even scholarly articles. Many maps of the Pacific that are frequently used have severe problems. Most often, researchers who work on the Pacific use decent maps, but if you look at non-humanist scholars, newspaper articles or text books the images found there are often disappointing. In this project, we try to rectify that.

For example, above is a very, very simplified world map that shows major land continents and barriers between them. It was uploaded 2018-11-03 to the facebook group "I feel personally attacked by this map" by Dominic HT, (original source unknown at time of writing). If the aim here is to show which major barriers separate these landmasses and their rough geographic relation to each other, then this is a map well done. Needless to say, it lacks in many other regards and shouldn't be used to illustrate anything else. For example, we certainly would not like to use this map to illustrate the pacific ocean and the islands therein.

While it is true that maps vary depending on their aims; there are some commonalities between maps that illustrate the Pacific. When constructing such a map, it's worth avoiding these 4 problems:

1. non-Pacific centered maps - this is obvious (see map 1)
2. only show landmasses and no island groupings. Granted, this depends on what you want to show, but chances are you're almost always better of somehow showing which islands should be grouped together by some kind of polygons, probably EEZs (see map 2)
   • if you do have some kind of groupings of islands, make sure it's clear to the reader what those grouping are based on (see map 4)
3. maps that leave out some islands of the Pacific (if you only focus on the West Pacific, don't call it a map of the Pacific, that implies the whole) (see map 3)
4. straight lines representing maritime boundaries from various treaties, "boxes" (see map 3)
   • Different treaties mean different things, i.e. not every line on such a map is the same. Each treaty has different implications. If you don't intend to discuss and make use of those maritime boundaries, choose something simler and more coherent - like the Law of the sea that the United Nations Convention decided on in 1982. Exclusive Economic Zones (see more on EEZ below) are a part of that convention, and much simpler and easier to show and relate to. Chances are, EEZ are better for your map than the square treaty boundaries.

Now that we've had a little jab at the current state of Pacific maps, let's move forward with some better maps. The aim of the maps under construction here are to:

• show the interconnectedness of the islands and nations of the Pacific Ocean, both pre-colonially and today
• provide the general public and educational institutions (especially in the Pacific) with more accurate and consistent illustrations of the region
• show the contemporary political situation in the Pacific and make people more aware of the French, American, New Zealand, Indonesian and Chilean contemporary rule in the Pacific
• illustrate the distribution of languages in the Pacific

Classical divisions of the Pacific

Exclusive Economic Zones (EEZs)

Ref: CartoGIS, ANU College of Asia and the Pacific, The Australian National University. This image is under Creative Commons Copyright (BY = attribution & SA = Share Alike). 2017

Hilario De Sousa was helpful in pointing out small errors in earlier versions of this map (Kingman and Jarvis were marked as Kiribati).

This map shows the Exclusive Economic Zones (EEZs) of the Pacific, i.e. the territories over which states have special rights to extract resources. It is not the same as territorial waters. In this case, this mainly relates to the fishing industry. For more information on EEZ, please read the wiki-article. This map is helpful because it visualises bordering states more clearly, the pacific looks less like a vast area with a scattering of land masses and more like a connected space (which it is). EEZs are 200 nautical miles from the coast. As a point of comparison, Jeff Marck argues that a reasonable overnight voyage in pre-colonial times would have been 100 nautical miles. He has also made maps of this in the pacific to show these polygons, more on those below.

Notes

• The Matthew & Hunter Islands are contested territory, they have been claimed both by Vanuatu and France (New Caledonia)
• New Caledonia will soon hold another vote on whether they should become independent or not, which would significantly change this map
• Several of the areas held by the USA in the pacific consist of entirely or almost uninhabited islands/atolls (Jarvis, Kingman, Howland & Baker and Wake)
• The EEZs have been coloured according to the "top-political unit" (not the same as sovereign state), for example: Tokelau is a dependency of the New Zealand
• possibility: colour different kinds of governance among dependencies differently (New Caledonia - French Polynesian, Cook Islands & Niue - Tokelau etc)
• Only political units that have at least one island in Oceania have been coloured in, i.e. Philippines and Mexico are not highlighted.

Planned adjustments

• grey out the areas that are not in Oceania so that it's cleared why the "south China sea", Mexico etc aren't coloured in. This would include non-New Guinea Indonesia, christmas island etc
• "happier" color for Federated States of Micronesia
• mark out national capital cities consistently, and not any other cities (i.e. remove Darwin etc)

Ref: CartoGIS, ANU College of Asia and the Pacific, The Australian National University. This image is under Creative Commons Copyright (BY = attribution & SA = Share Alike). 2017

Hilario De Sousa was helpful in pointing out small errors in earlier versions of this map (Kingman and Jarvis were marked as Kiribati).

This map shows the Exclusive Economic Zones (EEZs) of the Pacific, i.e. the territories over which states have special rights to extract resources. It is not the same as territorial waters. In this case, this mainly relates to the fishing industry. For more information on EEZ, please read the wiki-article. This map is helpful because it visualises bordering states more clearly, the pacific looks less like a vast area with a scattering of land masses and more like a connected space (which it is). EEZs are 200 nautical miles from the coast. As a point of comparison, Jeff Marck argues that a reasonable overnight voyage in pre-colonial times would have been 100 nautical miles. He has also made maps of this in the pacific to show these polygons, more on those below.

Notes

• The Matthew & Hunter Islands are contested territory, they have been claimed both by Vanuatu and France (New Caledonia)
• New Caledonia will soon hold another vote on whether they should become independent or not, which would significantly change this map
• Several of the areas held by the USA in the pacific consist of entirely or almost uninhabited islands/atolls (Jarvis, Kingman, Howland & Baker and Wake)
• The EEZs have been coloured according to the "top-political unit" (not the same as sovereign state), for example: Tokelau is a dependency of the Realm of New Zealand
• Only political units that have at least one island in Oceania have been coloured in, i.e. Philippines and Mexico are not highlighted.

Planned adjustments

• grey out the areas that are not in Oceania so that it's cleared why the "south China sea", Mexico etc aren't coloured in
• "happier" color for Federated States of Micronesia

Subgroupings of Oceania: Near/Remote Oceania v.s. Melanesia/Polynesia/Micronesia

CartoGIS, ANU College of Asia and the Pacific, The Australian National University. This image is under Creative Commons Copyright (BY = attribution & SA = Share Alike)

This map shows the regions of Polynesia/Micronesia/Melanesia while at the same time also the regions of Near Oceania and Remote Oceania. Both of these subdivisions are useful for anthropologists, linguists, archologists and other researchers of Pacific history.

This map shows both at the same time, it highlights how different classifications are useful for different approaches. The South Pacific can be divided up in many different ways, Polynesia/Micronesia/Melanesia is an old division that dates back to western colonisers in the 17-1800's. Naturally, the people who lived there then didn't divide up the world in this manner.

Antoinette Schapper pointed out to me over Facebook that the western borders of Melanesia could be redrawn to include more of eastern Indonesia. In this case, we went with older maps as a reference, and chose not to redraw those borders. It's a very worthwhile point though. If you do use this map in a publication or teaching where you focus on western Melanesia or Near Oceania, it may be good to underline this.

Notes

• Near Oceania/Melanesia could include more islands or easter Indonesia
• Macquarie island (South of New Zealand) was uninhabited when western seal fur traders reached it, but they did find ship wrecks of "ancient design". It is possible that Polynesian explorers did reach that far, or that at least they were aware of the island. Macquarie Island is also included in the Antarctic ecological subregion.
• There are societies of western Micronesia that most likely did not arrive there from the east (Polynesia/Melanesia), but from the west
• All languages spoken in Remote Oceania prior to the arrival of colonial white people were austronesian
• There are genealogically polynesian languages spoken in Melanesia, so called "Polynesian outliers". They are traditionally not highlighted in maps of Oceania subregions.

References

• Pawley, A.K., and R.C. Green, 1973. Dating the Dispersal of the Oceanic Languages. Oceanic Linguistics, 12:1- 67.
• Pawley, A. (2007). Locating Proto-Oceanic. In Ross, M., Pawley, A., and Osmond, M., editors, The lexicon of Proto Oceanic - The culture and environment of ancestral Oceanic society. Vol. 3: The physical environment, volume 3 of The lexicon of Proto Oceanic: The culture and environment of ancestral Oceanic society. ANU Electronic Press, Canberra, 2 edition.

Planned adjustments

• add in Australia as an Oceania region (different from Near and Remote Oceania)

Polynesian Outliers

Since the Polynesian societies in Melanesia, the so called "outliers" are often overlooked and aren't shown on maps of the subregions, here's two maps showing them.

When we divide up the world, we do so with different ideas of what it is that is interesting to highlight. Most maps of the pacific either show land masses with labels for countries and other places, Exclusive Economic Zones (see above) or larger regions (Polynesia, Remote Oceania etc). Anyone who has studied the Pacific or who is from the Pacific themselves knows that it is essential to take water into account, not as a barrier but as a road. Jeff Marck, a linguist who specialises on Polynesian languages, suggested a new way of dividing up water area in the pacific - overnight voyage distances by traditional canoes. If you are interested in pre-colonial history in the pacific, these voyage distances are more interesting than EEZs or land masses.

These maps show buffer zones ("blobs") around land masses that stretch out to 100 (land) miles. If two islands are within each others zones, the blobs are joined. 100 miles is Marck's estimate of a reaonsbale overnight voyage. Note that these areas can form chains where the ends of the chain aren't within 100 miles of each other.

The idea is that if two communities are within an overnight voyage of each other, they are likely to be in significantly more contact and therefore share more linguistic and cultural material. For Micronesia and Polynesia, these areas of overnight voyages corresponded well to linguistic boundaries . In Fiji there are more languages within these areas than there are within similar areas in Polynesia. As we travel further west, to New Caledonia and Vanuatu, we get even more languages per area (and per island). It would seem that Melanesian societies, compared to Polynesian, are more likely to develop language rich islands and island groups.

If two islands are more than 100 nautical miles apart, it does not mean that they were not in contact. The suggestion is that they were probably in less frequent contact than those closer to each other. It is clear that austronesian voyagers were capable and willing to travel much farther than 100 nautical miles, they discovered basically all islands of the entire pacific and possibly even visited south america.

These areas of more frequent contact show us a picture of the pacific which is a bit different from the one people are used to, they show an interconnected region of the world where the sea is not a divider, but rather a highway of transportation. Epeli Hau'ofa wrote about viewing the pacific as a "sea of islands" instead of a desert of isolated communities, and I believe that is a fitting description of what these maps aim to show.

The following maps are all based on this idea of overnight voyages by Jeff Marck. I start with showing the maps that he produced during his time at ANU.

Marck's maps

Micronesia

Eastern Polynesia

Western Polynesia + Fiji & Rotuma

These three maps of Micronesia, Eastern Polynesia and Western Polynesia + Fiji & Rotuma were all made by Cartography at the College of Asia-Pacific at the Australian National University, for Jeff Marck. The map of Micronesia is from a publication from 1986, the other two are from 2000.

Please note that there are some errors on these older maps, the Namu atoll and Jabat island for example are missing from the Micronesian map. These maps are not under revision, submitting comments on them to me is not relevant.

References

• Marck, Jeff 1986 Micronesian dialects and the overnight voyage. Journal of the Polynesian Society 95(2):253-258 (pdf)
• Marck, Jeff 2000 Topics in Polynesian Language and Culture History. Canberra: Pacific Linguistics

Note

• The overnight voyages are estimated to be 100 miles (c.f. 200 miles of the EEZs above)
• These estimates do not take into account winds, currents, reefs or other important variables for sea-faring

Below are the same maps, but with languages labels from Glottolog added on by me.

Hedvig's maps of overnight voyages

This is the most recent version of the entire pacific based on Marck's principles of overnight voyages. This map was made in 2018 at ANU CartoGIS, and will be finalised in 2019. This map serves as the backdrop of other maps that show the distribution of languages.

This map is different from Marck's not only in scope, but it also has non-western places as often as possible. I.e. if there exists an indigenous place name, that has been premiered over western names.

Uninhabited and uninhabitable atolls and islands are all marked out and named since they are all part of the pacific world.

Principles behind the drawing of contact areas and other technical notes

The way to read this maps is that if two islands are within 100 nautical miles or less of each other, their polygons are joined (polygon = "blob"). These polygons can be linked as chains, where one island at one end is only indirectly linked to the an island at the other end.

This map is based on the coastline data developed and maintained by Paul Wessel, SOEST, University of Hawai'i, Honolulu: http://www.soest.hawaii.edu/wessel/gshhg/ . The projection is azimuthal equidistant and the the central meridian is 180º. Staff at ANU CartoGIS used ArcGIS mapping software to draw 50 nautical mile rings around the islands, then these were “dissolved” – i.e. where they overlap, they get joined into one item/one polygon. Then another 50 nautical miles buffer was added. Where there is overlap of the polygons, it means that the two contact areas are overlapping in the ocean, but that they do not cover the same landmasses.

Please note that 100 nautical miles is an approximation of voyaging distances. We have not taken winds and currents into account here. Wind and currents, obviously, significantly impacts travelling distances. For an account of contact patterns based on drift, see:

• Levison, M., Ward, R. G., and Webb, J. W. (1973). The Settlement of Polynesia - A Computer Simulation. Australian National University press, Canberra.

Naming principles

For this map, we are using names that are not derived from western names or words to as large extent as possible. The reasons for this are twofold:

• the aim is to show the inter-connectedness of pacific islands prior to european influence, it is more appropriate to use non-western names
• the regions that emerge from this method don't always neatly map onto the categories the western place names denote.

Please note that the naming used on this map may not be what people living in that area use themselves and it should not be seen as a political statement of what things "should" be called. The intention is to give a view of how the pacific world was connected prior to european arrival. If you are interested in maps of the contemporary pacific and nation states, please see the EEZ (Exclusive Economic Zones) map here. That map shows the modern nation states and the water where they can exploit resources and police commercial usage by others.

The main focus of this map is Remote Oceania (i.e. Micronesia + Polynesia + Vanuatu + New Caledonia + Fiji + Reef and Santa Cruz Islands). New Guinea, Solomons and Australia have received less detailed attention. This is because the aim is primarily to illustrate contact between islands of Remote Oceania since this is the topic of my PhD work. Comments about Near Oceania are still welcome though, it is just I don't have the expertise or resources to address all those issues on my own.

This map contains place name labels that firstly correspond to the whole polygon. If there is no name that neatly refers to the group of islands, the names of smaller entities within the group are labeled. For example, the Samoan archipelago is labeled as "Sāmoa" because that name neatly does cover the polygon that emerges from the overnight voyages and it is an appropriate local indigenous name. The individual islands of Savai'i, Manono etc are not marked out since they are subsumed under the label "Sāmoa". In contrast, the two islands Rakahanga and Manihiki are both marked out separately even though they are within the same polygon, because there is no one term that covers the group as a whole.

Principles of naming:

a) is there an indigenous name that maps onto the entire 100 nautical mile polygon?

b) are there indigenous names for smaller groups of islands or atolls of the polygon?

c) are there indigenous names for islands or atolls of the polygon?

d) is there a western-derived name for the whole polygon or parts of it?

If there is (a), then the smaller entities are not marked out. For example, 'Upolu and Savai'i are islands within Sāmoa but they are not marked out because "Sāmoa" already covers them. This is not meant to be a detailed map of every island of the pacific labeled with its indigenous name.

Comments that do not abide by these principles will not be implemented.

Caveat regarding naming

It is very difficult to always use the most authentic place names, this map shows what currently is known to be the best non-western name for the polygon. Updates are always welcome.

As Lorenz Gonschor wrote to me, this map includes:

1) "indigenizations" of Western names

2 ) newly invented names in indigenous languages for islands where the knowledge of pre-European names is not precise

3) the islands were settled in colonial or postcolonial time by Pacific peoples from a language group other than their hypothetical pre-European discoverers and those new people renamed the islands in their native language

4) a European "discoverer" gave them names in another indigenous language from somewhere else.

This is regrettable, and updates that make this map better are appreciated. Ideally, the names would all be from similar time periods (for example, the 11th century) and reflect the actual name that people of the area themselves used. This however is very hard and sometimes impossible. In the meantime, please be patient and do not use this map as a definite map of indigenous place names. It is not meant to be that, it is meant to show these contact areas of 100 nautical miles and give an idea of interconnectedness prior to western colonization. In order to highlight the aim of the map, we've chosen to use non-western names. The alternative would be to continue using "Kiribati", "Marshall Islands", "French Polynesia", "Northern Marianas" etc, I'm sure you can see how that would be distracting from the pre-colonial focus.

If anyone else has the resources to launch a research project that seriously looks at the history of the pacific and naming practices, I'd be delighted. Let me know if that is the case!

CALL TO ACTION - Anthropologists and historians: Update wikipedia!

Pacific history is currently underrepresented on wikipedia. As I have been looking for information for labelling of island groups in the pacific, I've encountered some fascinating stories and some depressing ones. There are many cases where the wikipedia article on smaller atolls and island groups start right away with western colonisers, H-bombs and guano-mining with no mention of how the island or atolls fits into the landscape of austronesian voyaging or other historic events. This is unfortunate, and feeds into the narrative of the pacific as a barren empty space that was filled by western ships. In linguistics, there is an initiative to try and improve language pages on wikipedia - lingwiki. If anyone reading this teaches pacific history, I highly recommend making it an assignment to contribute to Wikipedia to improve this situation (under your supervision).

Specific areas where feedback is appreciated

As I go over revisions of the map and look up more info, these are the areas where there is still uncertainty. Comments on other areas of the pacific are also welcome, but these are the regions where advice is most needed

• The use of the name “Otuho” for what is now known as “Tepoto” in the group known as “Disappointment Islands” in French Polynesia
• Other names for “Suwarrow” in the Cook islands
• Distinguishing out the specific islands of the “Duke of Gloucester Islands“
• Naming practices in the 1988 ecological survey of the Caroline Island (Kepler & Kepler)
• Other names of the Matthew and Hunter Islands
• Other names of Norfolk island
• Other names of Chesterfield islands
• Names of archipelago consisting of of Satawal, Woleai, Lomotrek, Gaferut etc (if there is indeed one)
• Teasing out Japanese islands in the pacific (Bonin etc)
• Other names of the Line Islands (Kingman Reef, Palmyra, Jarvis, Kirimati etc)
• Teasing out groupings in Southern Cook Islands
• Names for Bounty and Antipodes islands
• Names for Phoenix islands

Planned adjustments

• grey out the areas that are not in Oceania so that it's cleared why the "south China sea", Australia etc don't have labels
• update place names according to most comments below
• remove reefs that are impossible to stay overnight on

Overnight voyages + languages

This map is based on the preceding map and shows the location of languages, based on Glottolog. The scope is restricted to Remote Oceania (i.e. Micronesia + Polynesia + Vanuatu + New Caledonia + Fiji + Reef and Santa Cruz Islands). The rest of Melanesia (rest of Solomons + New Guinea) and Australia are excluded here since the aim is to show the contrast between the different areas of Remote Oceania. I do not have the resources or expertise to re-examine all the names and exact locations of all languages of New Guinea and Australia.

The current version of the map is under more intense revision than the map above, since there are yet more names. It is possible to extend the scope of this map to "Oceanic Austronesia", i.e. the area of the Oceania where Austronesian languages are spoken. This would include parts of New Guinea and all of Solomons. At this time, it has not been decided if the map should extend further westward. Comments regarding this are also appreciated.

The aim of this map is primarily to illustrate the difference in language diversification in Remote Oceania, with Vanuatu sporting a whooping 135+ languages and Polynesian islands being more prone to one group = one island. This is quite nicely shown if you take into accounts Marck's 100 nautical mile polygons and map languages on top. This is part of my PhD project, which I'm carrying out here at ANU under the supervision of Andy Pawley, Nick Evans, Mark Ellison & Simon Greenhill. I'm exploring a lot of different factors in relation to this difference in diversification, we can discuss that separately as well if anyone is interested.

Principles of language classification (dialect vs language), naming and location

The data on languages for this map are taken from Glottolog. You can see their principles of naming, location and classification here.

Many requests for changes in this map are also changes that should be considered for Glottolog itself. I'm not keen on maintaining a separate database of languages to Glottolog, I'd prefer it if Glottolog was as good as it can be and the discrepancies between my data and Glottolog was kept to a minimum. I'll be evaluating the comments and if they are indeed things that should be changed in Glottolog I will submit them there. Alternatively, you can also submit change requests directly to Glottolog. You can either submit change requests as Github issues for Glottolog, or email them by clicking the orange alarm bell symbol on the relevant glottolog page for the languoid. If the Glottolog editors agree with our requests for changes, they will be implemented in the next version. If not, there may be discrepancies.

It is greatly appreciated if you consider the Glottolog principles and log a request there prior to contacting me.

Planned adjustments

• grey out the areas that are not in Remote Oceania so that it's cleared why the "south China sea", Australia, PNG etc don't have labels
• make separate maps for western polynesia + fiji + rotuma, eastern polynesia, micronesia, vanuatu and new caledonia
• Only have language labels for languages in Remote Oceania
• make it so that the non-insert versions of Vanuatu, Fiji and New Caledonia have small red dots for the language, but not the name labels. Name labels should only be in the insets.
• update various language names (see extensive comments below)

Below is the same background map again, but with languages as dots coloured according to language family. In this map all languages of Oceania (including Australia) are represented.

Planned adjustments

• grey out the areas that are not in Oceania so that it's cleared why Sulawesi etc don't have points
• change color of Pama-Nyungan to purple instead of red

NEXT:

Language history mapped to geography

What's next you ask? Well, wouldn't it be neat to map language family tree relationships of these languages onto geographic locations, i.e. on maps? Yes, yes it would. Turns out, as is often the case, that if you have a good idea - chances are someone else has already had that idea. In this case, we've got at least two cases of this for Austronesian (Kikusawa and Levison, Ward & Webb) and one very famous example for Indo-European. I'll show case all three of them here.

You should always keep in mind that the history of languages is not necessarily the same as the history of the speakers. People consist of genes, language consist of memes. These are not subject to the same pressures and mechanisms, and can develop independently.

In all of these examples, each language, and also each proto-language, is represented by one geographic dot and the dots are joined by the proposed trees for the family/branch. Languages are spoken over areas, not "dots", but this is a simplification that is done in this case in order to relate to the trees.

Indo-European (Bouckaert et al 2012)

This is a video showing the results from a study of the history of Indo-European languages, by Bouckaert et al 2012 in Science. It shows the expansion of the family given their study, from Anatolia to India, and also throughout Europe.

References

• Remco Bouckaert, Philippe Lemey, Michael Dunn, Simon J. Greenhill, Alexander V. Alekseyenko, Alexei J. Drummond, Russell D. Gray, Marc A. Suchard & Quentin D. Atkinson (2012) Mapping the Origins and Expansion of the Indo-European Language Family. Science.

Austronesian (Kikusawa, National Museum of Ethnography, Japan)

This map is by Ritsuko Kikusawa of the National Museum of Ethnography, Japan. It shows the expansion of the Austronesian language family, as defined by Crowley et al (2011) and Adelaar & Himmelmann (20119. Supplementary data was also taken from Ethnologue.

References

• Crowley, Terry, John Lynch and Malcolm D Ross. 2011. The Oceanic Languages (Routledge Language Family Series) 1st Edition. Routledge.
• Adelaar, K. Alexander and Nikolaus Himmelmann (eds). 2011. The Austronesian Languages of Asia and Madagascar (Routledge Language Family Series) 1st Edition. Routledge.

Polynesian settlement (Levison, Ward & Webb (1973))

This map is from a book on computational models of the settlement of the pacific (Levison, Ward & Webb 1973).

The boxes represent proto-languages, i.e. a shared language that we think was spoken by communities that now speak two different languages. For example, linguists believe that the people that went on to settle Tonga and Samoa used to be one population and that they spoke one language, we call this language "Proto-Polynesian". The placement of these boxes of proto-languages is not particularly meaningful, it's somewhere between the actual islands to indicate that there was something shared before settling or that they maintained a shared language in the different islands for a while after having settled.

This language history tree was based on knowledge at the time, it has since been revised. Below is a modified version of the map where I have updated it to more contemporary trees of Polynesian language history.

References

• Levison, M., Ward, R. G., and Webb, J. W. (1973). The Settlement of Polynesia - A Computer Simulation. Australian National University press, Canberra.

Notes on Levison, Ward & Webb (1973)

• The abbreviation PT occurs twice, once for "Proto-Tongan" and once for "Proto-Tahitian". In the map, the box of PT to the left is "Proto-Tongan" and the one to the right is "Proto-Tahitian"
• The fact that "PS" stands for "Proto-Samoan" was not on the map in the printed version, someone has written that in by hand in the copy at the Australian National University library
• Language history and population history (genetic history) are not the same, it may be that people were speaking a different language in the first settlement period and then due to contact or dominance from other groups, they changed language. This is always important to have in mind when looking at language history for clues of population history
• Pacific islands were not isolated from each other, people frequently travelled and most likely maintained dialect chains for a long time before "breaking apart" into different languages. When studying pacific history, this is always important to keep in mind. It is not the case that people migrated to a new island group and then stayed put and had no further contact with the islands the originated from and the islands that were later discovered

NEXT (2):

The prettiest AUSTRONESIAN tree

Sometime, it would be fun to produce a beautiful illustration of the history of Austronesian languages in the shape of a fishtail palm tree - simuarlly to how Minna Sundberg illustrated the Indo-European and Uralic families using oaks.

References

• Gray, R.D., Drummond, A.J., & Greenhill, S.J. (2009) Language Phylogenies Reveal Expansion Pulses and Pauses in Pacific Settlement. Science, 323: 479-483.

You can submit comments and suggestions here for the maps that are under development.

Below are the edits that have been suggested so far, that aren't listed under "planned adjustments" already. Please check if your query has already been raised before submitting.

Tepoto (Hedvig)

There’s two tepoto. One is within the Tuamotu polygon and one is next to Napuka.

Tepoto (north): 14°08′S141°24′W

Napuka 14°10′12″S141°13′52″W

Tepoto (south): 16°48′S 144°16′W

Only Napuka & Tepoto North need to be marked out, Tepoto south shouldn't be marked

Marquesas small change (Hedvig & Emily Donaldson)

The two names are “Te Henua 'Enana” & “Te Fenua 'Enata”

Te Fenua ‘Enana -> Te Fenua 'Enata

Mapia (Hedvig)

There isn’t more buffer needed for Mapia?

Hunter & Matthew (Hedvig)

We don’t need to label the western names too, the Oceania names are enough. Remove “Hunter” and “Matthew”

Mokil (Takuya Nagaoka)

Mokilese -> Mwoakillese

Tobi (Don Rubinstein)

"Tobi" is an anglicized version of an abbreviated name for Hatohobei. Hatohobei is now the recognized spelling of the island (and recognized as one of the 16 modern states comprising the Republic of Palau).

Tobi -> Hathobei

Yap (Don Rubinstein)

The indigenous name for Yap is Wa'ab (sometimes found as Waqab, depending on how the glottal stop is represented in orthography).

Yap -> Wa'ab

Wallisian (east uvean) - > Fakauvean (Léuli Eshraghi)

West uvean -> Fagauvean

Guam and Marianas (Don Rubinstein)

According to some early 17th century Spanish sources, the large southern islands of the Marianas, including Guam, Saipan, and Rota, were known collectively as Laguas; and the small islands of the northern Marianas, including Pagan, were known as Gani. There is now some interest among Chamorros in reclaiming these terms, although they're not well known among contemporary islanders, and they don't appear (or only very rarely appear) on early European maps . An Internet search of "Gani Islands" will turn up a few relevant sites.

Guam + Saipan + Rota -> Laguas

Pagan -> Gani

Sorol (Eric Metzgar)

I’m not clear on your rationale for continuing to label Sorol Sorol. Fais is an intermediary island approx. 46 nm from Ulithi and 98 nm from Sorol. This puts Sorol within the 100 nautical mile principle within voyaging range of Fais, and thus within the “Ulithi” sphere of influence.

Unicode representation

Something went wrong in CartoGIS with character encoding. Lots of language names in Vanuatu and New Caleodonia need to be ammended.

Temotu language (Alex F)

in Temotu, "Vano" is assigned to the wrong island; it belong in Vanikoro

Howland, Baker and Jarvis (Lorenz Gonschor)

the three islands were temporarily inhabited by Hawaiian guano miners who gave them the following names: Howland = Ulukou; Baker = Puaka'ilima; Jarvis = Paukeaho.

Ontong Java (Hedvig & Oliver)

Change to Luangiua

Kirimati -> Kiritimati

Kanak languages (Nick Thieberger)

there are character encoding typos in languages of Kanak ("New Caledonia")

Society islands/toatiete mā names (Lorenz Gonschor)

Split into Ni'a Mata'i (Windward Islands, i.e. Tahiti, Mo'orea and two small uninhabited islands nearby) and Raro Mata'i (Leeward Islands, i.e. Ra'iātea, Taha'a, Huahine, Pora-Pora etc.).

Dehu -> Drehu

Terei -> Buin

Santo

Tutuba location is off

Make names less Western

remove suffixes "-ese" and "-an" everywhere

Kwamera (James Flexner)

Kwamera -> Nafé

Ambrym

remove Repanbitip

Add in coordinates for languages of Remote Oceania and Australia that are currently missing locations in Glottolog

Marianas & Guam (Ash Sablan and Travis B. Wells)

it might be better to label the entire archipelago, “Laguas yan Gåni” (as opposed to splitting it up into Laguas and Gåni respectively) because the entire chain is indeed populated by a single (Chamoru) nation.

Hawai'ian northern islands (Lorenz Gonschor)

An example of the second category are the Hawaiian names given on the map for the Northwestern Hawaiian Islands. Nīhoa is a authentic pre-European name and Mokumanamana likely as well, but beyond that the precise pre-Europena names are doubtful. There are Hawaiian traditional history sources mentioning names for them, and Hawaiian archaeologist Kekuewa Kikiloi has recently attempted to reconstruct which name fits which island, and I would recommend using his article www.ksbe.edu/_assets/spi/hulili/hulili_vol_6/5_Rebirth_of_an_Archipelago.pdf here, because his is the one best effort in that sense so far. Then there was a move during the Hawaiian Kingdom in the 19th century (i.e. post-contact but still under native control) during which two of the outer islands received Hawaiian names, Moku Pāpapa for what Europeans call Kure (the last in the Northwestern Hawaiian chain) and Kalama for what on Western maps is called Johnston (the little atoll to the southwest of the Hawaiian chain). The names used on the map here for the Northwestern Hawaiian chain, on the other hand, are inventions by a Hawaiian language commission in the 1990s and have little relation to serious efforts of reconstructing ancient names (like Kikiloi did) plus they conflict with the 19th century Hawaiian Kingdom naming efforts by assigning "Moku Pāpapa" to what Europeans call French Frigate Shoals instead of Kure.

The map below is based on Kikiloi's name classification, it's from the website of the governmental conservation area - Papahānaumokuākea Marine National Monument.

Acknowledgements

Thanks for advice and comments regarding this map goes to members of the ASAONET mailing list and Facebookers from the group "Maps of the Pacific".

Special thanks goes to

• Emily Donaldson
• Don Rubinstein
• Diego Muñoz
• Apo Aporosa
• Michael Goldsmith
• John Lynch
• Charles Langlas
• Hilário De Sousa
• Jonathan Schlossberg
• Sally Akevai Te Namu Nicholas
• Jeffrey Marck
• Myjolynne Marie Kim
• Alexander François
• Lorenz Gonschor
• Mary Walworth
• Marianne “Mimi" George
• Sylvia C Frain
• Matthew Spriggs
• Oliver Sheehan
• Joseph Watts
• Ritsuko Kikusawa
• Takuya Nagaoka
• Nick Thieberger
• Liam Kokaua
• James Flexner
• Tom Dougherty
• Ash Sablan
• Travis B. Wells
• Léuli Eshraghi
• Angela Cincotta-Segi