Previous Next

Aboriginal Placenames

14. Naming Bardi places

Claire Bowern


In all cultures, toponyms (or placenames) are important, for they form an integral part of our history and culture. When learning an Australian language, knowledge of the placenames is vital to becoming a fluent speaker. This is especially true of the Bardi language. In Bardi, toponyms are often used instead of relative directional words or deictic markers when giving directions or describing locations. They appear frequently in narratives; it is impossible to follow a conversation or story in Bardi without a good knowledge of the names of places.1 Older Bardi people also consider placenames to be among the most important aspects of their culture. During the planning for the Bardi dictionary, for example, the senior man who started the project was adamant that mythology and placenames were the most important things to record. Other old Bardi people have expressed concern that younger people no longer know many of the placenames and so cannot describe where they have been; the older people feel this lessens their connection to the land.

In this paper I will explore some topics associated with placenames in the Bardi language and in Bardi country. I look at the structure of naming organisation and argue that the analogy of ‘suburbs’ in a ‘city’ is useful to describe the relationship between booroo names (see ‘Booroo’ below) and area names in Bardi country. Then, I make some notes on social organisation and how it relates to place and topographical organisation. Next, I look at the language of the names themselves, their etymology and relationships to story and myth. Finally, I make some comments on current naming practices, including the inclusion of Bardi names on English topographic maps.

The organisation of places and placenames

Bardi traditional country comprises the tip of the Dampier Peninsula, to the north of Broome, and the offshore islands at the western end of the Buccaneer Archipelago. It includes the former Sunday Island Mission and the current communities of One Arm Point and Lombadina/Djarindjin, and the outstations surrounding these communities, as well as the tourist resort at Kooljiman (Cape Leveque). To date, the Bardi dictionary and dictionary supplement contain 535 different placenames, and we still have many blank spaces on the map. On 14 June 2005, the High Court of Australia granted Native Title to Bardi and Jawi people over traditional country on the Dampier Peninsula and the surrounding sea (although not the off-shore islands).

Bardi placenames fall into a number of different types. The types are arranged hierarchically, although not all levels in the hierarchy have a name in Bardi. This organisation can be thought of as somewhat parallel to addresses in English. When we write an address we give the street, the suburb or town, the state and the country. Streets are contained in suburbs, which are divisions within cities. Bardi places are similarly hierarchical – although there were no streets and towns in traditional times, there were districts, areas, and named places and landscape features within these larger areas. In this section I describe placenaming practices.


The broadest named type of placename in Bardi is the ‘area’. There seems to be no term for this in Bardi (nor in the other related languages of the area), although the areas themselves are named. Areas in the northern part of the Dampier Peninsula tend to be about a day’s walk from one side to the other. There are six main ‘areas’ where Bardi is spoken.2 The areas are shown on Figure 14.1.

The northern part of the Dampier Peninsula, north of Beagle Bay, is divided into four roughly equal quadrants (or ‘areas’); Goolarrgoon is the north-western area comprising the resort area of Cape Leveque and environs. Goowalgar is the area south of Goolarrgoon, as far as Pender Bay. An alternative name for this area in Robinson (1973: 103) is ‘Olonggon’, which means ‘in the south’. On the eastern side of the peninsula, there is Ardiyol,3 the area including the community of One Arm Point, and Baanarrad, centred around Galan (Skeleton Point).4These names can be etymologised as containing the compass points goolarr ‘west’, baani ‘(south-)east’ and ardi ‘north-east’. The -goon of ‘Goolarrgoon’ is the locative marker. I have no etymology for Goowalgar.

Figure 14.1: Area names where Bardi is spoken


The fifth ‘area’ is Iwany, which comprises the three islands together called ‘Sunday Island’ on English maps. It also includes the islands immediately to the north (such as Ralooraloo) and the islands/rocks in Sunday Strait. The sixth area is Iinalabooloo, the islands between Sunday Island and the mainland. Iwany has no recognisable etymology,5 but Iinalabooloo is from *iinalang-bulu ‘island-dweller’.6 Another term which may be an area name is Mayala, the name for the islands of the Buccaneer Archipelago east-north-east of Sunday Island.

Perhaps surprisingly, given the usual fuzziness of boundaries in Aboriginal naming systems, the area names seem to be clearly defined (although I do not know if this was also the case in pre-mission and pre-land claim days). Of course the island areas (Iwany and Iinalabooloo) have defined boundaries (the ‘areas’ do not appear to include the surrounding sea; Sunday Strait, for example, does not seem to be part of Iwany), but old people are also quite definite about the end of Goowalgar (the southern shore of Pender Bay) and the ‘border’ between Goowalgar and Goolarrgoon. This is in contrast to the locality names (see Foreign placenames), which are less categorically defined.

The ‘area’ terms are also relevant to a discussion of marriage patterns. In all cases where the booroo or area of each spouse is known, they come from different areas. Iwany people seem to have had a preference for spouses from Baanarrad or Ardiyol, and Goowalgar people tended to take their spouses from Goolarrgoon. Some Iwany men married Mayala women, particularly Oowini women from the islands off Wotjalum.


Within each area, there are a number of booroo. The word translates as ‘camp, ground, place’ and also as ‘time, tide’, as in ginyinggon booroo ‘in that place, at that time’.7 The group of people who are identified with a particular booroo can be roughly equated with the Yolŋu bäpurru or ‘family’ (for which see below). That is, the booroo was a patrilineal estate, a place which would be owned by a group which formed an important part of Bardi social organisation.

Booroo are geographically much smaller than ‘areas’; for example, there are at least 15 booroo in the Goolarrgoon area. The booroo names are often also island names; thus Jayirri is both an island and a booroo, as is Jalan. Other islands may contain several booroo, as for example Sunday Island (Iwany) does.

‘Owners’ of places are determined patrilineally, and thus fishing and camping rights to particular places are determined by who one’s father is. Bardi people also have certain rights in one’s ngiyalmoo, or mother’s country. In the older texts people (particularly men, but also women) are often called by their booroo names. The following names are found in the Laves’ corpus of Bardi from the late 1920s (Laves n.d.). In some cases the location of the booroo has been lost and is recorded only in the name. In other cases, such as Albooloogoon and Narrigoon, the booroo location is known.

  1. 1. Albooloogoonbooroo Gabinyoogoonbooroo Galooroonbooroo Gamilbooroo Goolijoonkoonbooroo Gooloorroonbooroo Ilordonbooroo Joorrbaanbooroo Ngiyalbooroo Ngoorrgoonbooroo Yalyarinybooroo Narrigoonbooroo

The late Nancy Isaac has described this practice as like using a ‘surname’ or alternative designation for a person, although it should be noted that nowhere in my texts are people called by more than one name. One’s booroo appears to be the primary point of identification for senior Bardi people. For example, older Bardi people tend to answer the question ‘where are you from’ with the booroo and perhaps the area, whereas younger Bardi people tend to give the community, such as ‘One Arm Point’ or ‘Djarrindjin’.

These days some booroo names are well known, but many have been lost. The booroo names for Sunday Island (Iwany) are only known to the oldest Bardi people. In this case, ‘Iwany’ is now used the way that ‘Jayirri’ is used, even though ‘Jayirri’ was a booroo/island name and ‘Iwany’ was an area term. That is, ‘Iwany’ is now treated like a booroo name rather than an area name. On the other hand, several booroo names, such as ‘Ngamoogoon’,8‘Gambarnan’ and ‘Boolgin’ are still frequently used.

A set of example booroo names for the Goolarrgoon area is given on Figure 14.2. For further information, see Robinson (1973) and Smith (1987).

Figure 14.2: Booroo names in the Goolarrgoon area


Locality names

Each booroo contains multiple specifically named places within it. These may refer to different topological features, such as tidal creeks (iidarra), rocks (goolboo), reefs (marnany), caves (gardin), hills (garrin) or small bays, or they may refer to a piece of land behind the shore. Occasionally they also refer to man-made features, such as the large permanent fish traps (mayoorroo) around One Arm Point. Locality names typically refer to a feature, but the name will also be applied to the ground surrounding that feature. Thus ‘Jaybimilj’ refers specifically to two submarine depressions (where fish congregate) in the tidal channel between Iwany and Oombinarr; however it also refers to the stretch of the passage where the depressions are located (the passage as a whole is called ‘Jaybi’). See Figure 14.3 for the location of these places. Most locality names are not analysable, although some brief comments are made below in ‘Etymology and lexical categorisation of placenames’.

There are few recorded locality names away from the coast line. I do not know if this is a result of loss of traditional knowledge9 associated with the shift to English, or if it is a relic of older naming practices. I mostly suspect the latter – in the Laves texts, for example, places inland are usually referred to by their booroo name, whereas more specific locality names are used for areas along the coast. Bardi people are primarily coastal dwellers and seemed to have seldom camped inland. This makes sense in an area where the sea is a far more fruitful resource than the land, and where there is no permanent surface water (cf. also Smith 1983, 1984, 1987).

While booroo names are unique within Bardi country, locality names are not. Thus there is only one booroo called ‘Albooloogoon’, but there are multiple places called ‘Mardaj’. Where disambiguation is necessary, the booroo name is given as well. Thus ‘Jayirri Mardaj’ is ‘the Mardaj on Jayirri’. ‘Iilon’ is another common name. The places named ‘Mardaj’ are all reefs, and places named ‘Iilon’ are all valleys slightly inland away from the coast.

It is tempting to think of the booroo names as ‘big names’ and the locality names as ‘small names’ (see Keen 1995) as in Yolŋu and several other Aboriginal areas. In such systems, the ‘big name’ places are focal sites; they refer to a specific topographic feature and the area around it, although how far the name extends from the focal site is not fixed. Such sites are also often important ceremonial sites and are riŋgitjmirri (in Dhuwala), or associated with exchange networks between specific groups. ‘Small name’ places refer to specific features. Bardi placename organisation differs from this in several ways. Firstly, most booroo names do not appear to be focal, although some are; that is, they do not tend to refer to a specific feature and the area around it. Locality names have this characteristic, but booroo names do not. Secondly, booroo and locality names appear to be thought of hierarchically – thus a locality can be said to be ‘inside’ a booroo. This is not the case for small and big names in Arnhem Land. Small names can occur inside big named places, but they do not have to. Locality names appear to have the properties of both small and big names – thus their range may extend or contract depending on the knowledge of the speaker or how specific they wish to be. See Figure 14.4 for the non-secret locality names around One Arm Point (Ardiyooloon).

Figure 14.3: Some locality names on Sunday Island


Figure 14.4: Non-secret locality names around One Arm Point community (Ardiyooloon)


Multiple names for places

I have not encountered any places (booroo or localities) in Bardi country with more than one Indigenous name, although there are multiple instances these days of the Bardi name and the European name coexisting. The One Arm Point Community/Ardiyooloon is a good example. Gooljiman/Cape Leveque is another.

It has not been possible to do extensive research involving Bardi names for places outside Bardi country. I do not know to what extent Bardi people have their own names for places outside Bardi country and to what extent they use the names in the neighbouring language. In all examples I have, Bardi people use a name which is clearly not Bardi. The place Bolg, for example, which is close to Beagle Bay in Nyulnyul country, is known as ‘Bolg’ to both Nyulnyul and Bardi speakers (see Figure 14.5).10 The same is true for placenames I have recorded in Worrorra-speaking areas, as far as I have been able to substantiate (for example, ‘Barndijin’, in the area of Kunmunya Mission).

Language names and placenames

A comment is warranted on the relationship between area names and language names. The terms for the languages of the Dampier Peninsula are all Indigenous names. The names are constant from language to language and have no etymology. This is in contrast to some parts of Australia, where languages (and ethnic groups) are known by their words for ‘this’ (e.g. Dhuwal, Dhuwala, Djinaŋ and Djaŋu in Arnhem Land) or ‘no’ (such as Waŋaaybuwan and Gamilaraay in New South Wales).

The relationship between language names and placenames is not straightforward. Language and land are inextricably linked through creation stories; the same culture hero who named the places and showed people water on the islands also told people what language they speak. The text is given in examples (2) and (3) in the next section. The places are named in a particular language, and people from a particular place speak a particular language. Thus one finds statements that ‘Goolarrgoon people speak Bardi’ or ‘Goolarrgoon people are Bardi’ but not that ‘the Bardi language belongs to this area’. Thus language names and area terms seem to be to some extent independent. For example, although Goowalgar is a Bardi area, the Bardi of Goowalgar speakers is often described as ‘Bardi coming up Nyulnyul’ or ‘Bardi mixed with Nyulnyul’. Language terms also have much less clear boundaries than the area terms do.

Other names for places

A few other terms are also in frequent use. Gaanyga ‘mainland’ always refers to the Bardi-owned area of the Dampier Peninsula, not the ‘mainland’ in general (as opposed to an island). Nyanbooroonony ‘the other side’ always denotes the Wotjalum area on the opposite (eastern) side of King Sound from the Dampier Peninsula. This would seem to be similar to the use of ‘the hill’ in US English to refer to the Capitol in Washington, DC, or ‘the lake’ in Canberra to refer to Lake Burley Griffin (never, for example, Lake Tuggeranong or Lake George).

Etymology and lexical categorisation of placenames

Having described the system of placename organisation in Bardi, I now turn to the etymologies of the placenames in the language. Most booroo terms seem to have no etymology. A few contain the locative marker -goon (e.g. Ngamoogoon, Ngarrigoon) but the roots have no known etymology. The following comments apply mostly to locality names.

Descriptive names

Some locality names are clearly relatable to Bardi common nouns. ‘Laanyji’ ‘seaweed’ is one, an area of thick seaweed on the south-eastern part of Sunday Island, near the channel between Oombinarr and Iwany (see Figure 14.3). There is also a locality with the name ‘Noomool’, ‘seaweed’. Boolngoorr, a peninsula on the northern side of Jayirri island, is homophonous with boolngoorr(oo) ‘middle’, and it is between two other peninsula (see Figure 14.5).

Figure 14.5: Examples of descriptive locality names


In addition to the placenames where the name is obviously descriptive, there are many Bardi placenames which are relatable to common nouns but where the significance of the name is not apparent. ‘Aalinan’ (also known as Gilbert Rock) may be related to aalin ‘fish hawk’ (the suffix -an is the Jawi purposive suffix, cognate with Bardi -ngan). A few islands are paired into boordiji ‘big’ and moorrooloo ‘small’, such as Boordiji Ngaja and Moorrooloo Ngaja (‘Big Ngaja’ and ‘Little Ngaja’ respectively) in Mayala.

While many names are topographically descriptive, there is little evidence, if any, for a classificatory system of the type frequently found in European placenaming (as pointed out for other Australian languages by Hercus, Hodges and Simpson 2001: 15). A few placenames involve the word nankarra ‘forehead’, which is also used for promontories. There is also a certain amount of body-part/topographic feature polysemy (e.g. ‘ridge’ = ‘back’ (both niya)), but this body-part polysemy is not nearly as extensive as in some other languages (such as in the Yolŋu subgroup). Locality names do not seem to be classified into, for example, ‘mountain’ names (compare English Mount Hagen, Mount Ainslie), ‘island’ names (Treasure Island, Bathurst Island, Long Island) or river names (the Mississippi River, the Snowy River).

Placenames with -ngarr

These names are associated particularly with Worrorran-speaking areas. They are consciously associated with the islands in the Mayala group. For example, when we were trying to identify the places of names given in stories in the Laves corpus, on several occasions Jimmy Ejai and Bessie Ejai (my consultants) said that a place must be in an area “because it sounds that way, that language has -ngarr on its places”.

Placenames with -ngarr are not restricted to the islands shared with Worrorran areas, however, nor to the parts of Bardi country which strictly border it. They occur on Sunday Island (Iwany) and a few do occur on the mainland. ‘Jooljoolngarr’, for example, is a place on Sunday Island.

Foreign placenames

The islands in the Buccaneer archipelago (Mayala) are Yawijibaya and Unggarranggu11 country, but Bardi people have traditional rights of fishing, trochus gathering and access to various places. Some of the placenames in this country are clearly not of Bardi origin. Blog, for example, is the only word in the Bardi dictionary to contain an initial consonant cluster. The reef where Jooloo met the Moonyjangiid people is always called Brue Reef (or, occasionally by English speakers, Blue Reef, but Brue Reef seems to be the older, correct name). Bardi people say it is the English name, but no one has used the old Bardi name for it in living memory.

These days many placenames of English origin are also used by Bardi people. The area around One Arm Point (Ardiyooloon) contains several localities with both English and Bardi names, such as Factory/Bandilmarr, Hatchery/Algoonoomarr, Airstrip/Jimanygo, and Middle Beach/Galbarrnginy (see Figure 14.4).Further south, ‘Skeleton Point’ is more frequently used than the Bardi name ‘Mardnan’; the same is true for Cunningham Point/Garramal. However, some places are always called by their Bardi names, such as ‘Jologo beach’ (the swimming beach at One Arm Point). The English name ‘Middle Beach’ is only used by visitors to the community.

Names given by Galaloong and other culture heroes

Many placenames in Bardi country are said to have been placed there by the Bardi culture hero Galaloong. The relevant text (in a version told by the late David Wiggan to Gedda Aklif) is given below:

  1. 2. Jarri inanggalanan Galaloong boonyja booroo Nyoolnyool Banyiyola, nyalabalboora nirirr injoonoo. Booroo injoombarna irrnga, inamana irrnga booroo ginyingg aamba, irrngirrngi arrooloongan booroo barnanggarr. …Galaloong has been everywhere, Nyulnyul country and Bardi country, he’s been along the edge of everywhere. He’s been counting the names, this man ‘put’ the placenames, and we use those names now. (1-2)
  2. 3. ‘Ginyinggamba joogarra ngaanka Bardi.’ Ginyinggon barnin injoonoojirr, ‘barnin arraman barnimorr. Oola joogarra jiibi inin, biidini joogarra oola niiminiimi, joogarra oola,’ injoonoojirr.‘This is your language, Bardi,’ (he told them). Then he told them, ‘do things this way. Here is your water, your rockholes, your soakages in the ground and in the sand,’ he told them. (10-11)

A set of other (gender-restricted) names for certain localities was given by another culture hero. Some discussion of this topic can be found in Worms (1949, 1952, 1957).

Placenames and songs

Unlike some of the better-known Aboriginal culture areas (e.g. Arandic, Arabana, etc.), Bardi placenames are seldom overtly associated with a particular song or narrative story, or if they are, it does not form part of the name. For example, the southernmost bay on Iwany, Balalagoon, is strongly associated with a mythological raiding party of people from a Worrorran-speaking island (identified only as part of the Mayala group) who attacked people camped at Jooljoolngarr but who turned into dugongs when fleeing. The place has strong associations with the story – for example, the place is frequently described (in English) as “Balalagoon, you know, the place where those men turned into dugongs”. The place is not, however, named specifically from that story, and if Balalagoon means anything, it would be most likely related to balala ‘spinifex fishtrap’.

This is not to say, however, that Bardi people do not have song lines similar to those documented for northern and central Australia. Many song lines beginning in Worrorra and Ngarinyin country to the north-east of the Dampier Peninsula pass through the Dampier Peninsula. Some of these stories do not appear to be gender restricted in Worrorra, but they are in Bardi and are not further discussed here.12 Some localities have gender-restricted names and are associated with initiation rituals and with key events in the lives of culture heroes.

Placenames seldom appear in song language, even when the songs describe events located in specific areas (such as the events told in Billy Ah Choo’s ilma, or public song cycle). Even though ilma songs are set within a narrative framework, they are episodic in that each ilma verse describes an episode of the story in very abstract and metaphorical language. Prose interpretations of ilma verses always contain the places where the ilma verse is taking place, even though the ilma verses themselves almost never do. In the one case I have found of a placename in a song text, the song name for the place is different; it is ‘Balalbalal’, whereas the usual name is ‘Balalbalalngarr’. See Bowern (in preparation) for more details about Bardi song language.

Giving directions and describing places

In the previous section we considered the etymology of various placenames and their relationship to other parts of the Bardi lexicon. Let us now turn to the role of placenames in grammar.

Saying where something is

Placenames are used much more frequently than the deictic system (or absolute directions such as compass points) when giving directions or describing a journey. My inquiry on how to get to Lalaram from One Arm Point, for example, was answered by a list of the main places between Sunday Island and Lalaram. Compass points are seldom used either, although they do sometimes appear at the beginning and ends of texts. This is illustrated with the beginning of the story about Iidool, by the late David Wiggan, which starts with a woman travelling north and looking east and west for a ‘promise husband’ for her daughters. The placenames and direction terms are in bold:

  1. 4. 1: Nyoon inanggalanan gooyarranyarr bo. Biyambarr nyalabon ardi abarrabarr injoonan, arra oolalana aamba gorna aamba darrgal. Aamba oolalana arra.

A woman lived there [nyoon] with her two daughters. She wandered in the Biyambarr area, north-east [ardi], because she hadn’t found a proper man (for her daughters to marry). She didn’t see any men.

2: ‘Barda ngankayagoorr,’ injoonoojirr boyinirr.

‘I will take you both away,’ she said to her daughters.

3: ‘Goolarr ngankayagoorr barda baanarr ginyingg aamba nganggimi aamba balngarrinjoon joogarr aamba.’ injoonoojirr.‘I’ll take you west [goolarr] and away east [baanarr] I’ll look for this man, a promise husband for you.’ She said this to them.

4: Ginyinggo nyalaboo roowil ingirrinyan, nyalab garra garra garra garra, nyalaboo boor boonyja boor nyalaboo Garnmanyi jarri Barndijin.She walked this way [nyalaboo], she walked and walked and walked until she got to Kunmunya [Garnmanyi], on Barndijin Station.

In the first line, the narrator tells us that the woman wanders around in the Biyambarr area, and that it’s ardi ‘north-east’ of One Arm Point. Then in line three, the woman uses goolarr ‘west’ and baanarr ‘east’ as generic directions for where they will look for a husband of the appropriate marriage class for the daughters. Then in the next line the narrator uses the relative directional term nyalaboo ‘this way’ (towards the speaker) and two placenames, Garnmanyi ‘Kunmunya’ and Barndijin. The result is thus a combination of very generic directions (e.g. barda ‘away [from speaker]’) and specific locations, which are not very helpful for someone who does not already know where all these places are. Of course, until recently, anyone listening to the story would have already known the relative location of all of these places.

Use of English directionals

When older Bardi people give directions, they often use the following words, which are not used in the same way that they are used in standard English:

  1. 5. inside outside the other side this side

‘Inside’ refers to a place that is inland, or up the hill, from a coastal site. ‘Outside’ is the opposite, and is a place which is off-shore. It is not used for locality names within booroo names; for example, Mardaj is not said to be ‘inside’ Jayirri. ‘This side’ is always the mainland and surrounding islands of the Dampier Peninsula, while the ‘other side’ is always Yawijibaya/Umiida country. I assume that the English usage is a direct calque of Bardi nyanbooroonony ‘on the other side’.

Placenames in grammar

The word ninga ‘his/its name’13 can refer to both a person’s name and the name of a booroo or locality. In example (2) above, for example, the speaker tells us that Galaloong counted and put on the map the irrnga booroo, literally ‘their-names place[s]’. Placenames have a few features which set them apart from other nouns in the grammar.

Bardi placenames cannot productively bear the locative case -goon. The phrase ‘at One Arm Point [Ardiyooloon]’, for example, is translated as Ardiyooloon, not *Ardiyooloonkoon. Many placenames appear to have the locative case marker as part of the name, so the lack of locative marking may reflect an avoidance of double-inflection. Alternatively, since many words for topographic features (e.g. garrin ‘hill’, morr ‘road’) do not take locative inflection either, it may be a semantic restriction rather than a morphological one. Placenames can take allative and ablative inflection, which would further argue against the analysis of null locatives as a prohibition on double-inflection. Some examples are given below: (6) and (7) show zero locative inflection of a location noun and a placename respectively. (8) shows allative inflection on a placename.

  1. 6. Wayibalajininim jagoord ingirrini biila ginyingg morr (*ginyinggon morr).‘The group of white people returned on that path.’ (Metcalfe 1975: 83)
  2. 7. Niimana jarda aarli Ardiyooloon. Gorna aarli marlinngan.‘We have many fish at One Arm Point. They are good fish for eating.’
  3. 8. Arra marl oolirrgal roowil ingirrinyana barda garr joodinyko Nilagoonngan.‘Without stopping, they continued to Nilagoon.’

A locative marker -i has been recorded on a few placenames. It is only permitted on placenames, and not on other nouns:

  1. 9. Marroogoon, Galani nyoonoo, baarnka, Jibaji aarlon inanggalanana.‘At Marroogoon, at Galan there, outside, Jibaji was fishing.’

English names in Bardi country

When Europeans came to Bardi country and made maps of it, they used some of the Bardi names in addition to making up their own. The survey maps of Bardi show both European names (mostly English, and a few French) and attempts at spelling Bardi. There are several tourist maps of the Dampier Peninsula, although most also cover at least the Western Kimberley and give few placenames in Bardi country. Those that are given are always the English names (‘Curlew Bay’, ‘Thomas Bay’, ‘Goodenough Bay’). There are 1:250 000 scale maps of the whole of Bardi country, and 1:100 000 maps of parts, including One Arm Point and Sunday Island. In the following discussion, ‘the maps’ refer to Geoscience Australia’s 1:100 000 and 1:250 000 topographic maps.14 They were made between 1973 and 1983.

Names of European origin

Many of the bays and islands in Bardi country, and a few of the high points of land, were named after topographical features (e.g. One Arm Point, High Island) or after people (Dampier Peninsula, Cape Leveque) or their ships (Beagle Bay, Cygnet Bay). One Arm Point is so named because the peninsula is curved like the ‘arm’ of a semicircular bay; however, there is no corresponding arm at the opposite side of the bay. High Island rises steeply from the sea. The Dampier Peninsula is named after William Dampier, who landed in Bardi country in March 1689 to replenish supplies of water. Cygnet Bay, the bay where he stopped, is named after his ship.

Adaptations of Bardi names

Some names on the Geoscience Australia maps are clear English renditions of Indigenous Bardi names. ‘Poolngin’, for example, is clearly ‘Boolnginy’. The final ny in the Bardi orthography represents a palatal nasal, which has been missed in the English rendition of the name. ‘Alorra’ is ‘Ngolorron’, missing the initial ng or engma (a frequent change in Anglicising Indigenous names). ‘Lalowan’ is another adaptation, from ‘Laloogoon’.

A further few names are clearly Bardi, but refer to the wrong place. For example, ‘Tyra’ Island is clearly Bardi ‘Jayirri’, but the name on the Geoscience Australia map refers to ‘Manboorran’, the smaller island to the south. ‘Jayirri’ itself is given the name Jackson Island.

Two names, Tyra (= Jayirri) and Tallon (= Jalan) appear to show depalatalisation. The modern Bardi name has a palatal stop [j], while the name is recorded on the map with <t>. Some of the oldest wordlists have this correspondence in some words, and the Jawi dialect of Bardi also has this correspondence in some forms (cf. Jawi dool ~ Bardi jool, both ‘kneel’). The forms on the topographic maps could be evidence of a now-completed sound change which had not reached Jawi at the time the names were recorded, or they could be mishearings.

Ralooraloo, on the published maps as ‘Salural’, may be a typographical error. Bardi trills are often devoiced, and sound to speakers of European languages like an s, but devoicing would not occur in this position. It is most likely to be a typographical error, or a misreading of handwriting.

In one case, it is impossible to tell whether the name is Bardi or English. There is a resort at a place called Middle Lagoon, which in Bardi and Aboriginal English is pronounced [Miid(a)lagoon]. This name could be either the Aboriginal English pronunciation of ‘Middle Lagoon’, or ‘Middle Lagoon’ could be the English adaptation of an Indigenous placename ‘Miidalagoon’. Note that there is no lagoon in the vicinity, so if it is English the origin of the name is a puzzle. ‘Miidalagoon’ would be an unlenited form of ‘Miidaloon’, a name that occurs twice elsewhere in Bardi country. Another name where the direction of borrowing is uncertain is Chile Creek, which is Bardi ‘Jilirr’.

Assorted misunderstandings

Finally, some names in the area have come about through misunderstandings of the Bardi names for the places.

The former mission and now Community of Lombardina/Djarridjin (or Djarrinyin) is one example. It is variously spelled ‘Lombadina’ or ‘Lombardina’. The original name seems to have been based on ‘Loomarrdina’, the area where Jilirr creek enters the sea. At some point, however, it appears to have been etymologised as ‘Little Lombardy’.

Metcalfe (n.d.) records a place called ‘Oolarda’, given as the Bardi name for ‘Dish Island’. The Bardi name is actually ‘Diiji’ (and probably not related to English ‘dish’), which seems to have been later associated with English ‘dish’ and back-translated into Bardi as oolarda (coolamon). It may have been a pun amongst Bardi people when Metcalfe was compiling his dictionary.


Thus in conclusion, we have surveyed some of the issues involved in studying Bardi placenames. We have seen that toponyms are organised hierarchically, and that this type of organisation is rather different from that seen in some other, well-studied, parts of Indigenous Australia. We have also seen that Bardi placenames come from many sources, although not all of them are identifiable. We have also examined how toponyms form a distinct subclass of nouns in the language.


Aklif, Gedda 1999, Ardiyooloon Bardi Ngaanka: One Arm Point Bardi Dictionary, Kimberley Language Resource Centre, Western Australia, Halls Creek.

Bowern, C. (ed.) 2002, Jiiba nganman jawal: Stories in Bardi and English from the elders of One Arm Point and Sunday Island, transcriptions and translations of narratives collected under AIATSIS Grant G2001/6505, awarded to Claire Bowern and Bessie Ejai.

Bowern, C. (in preparation), ‘Bardi Song Language’, MS, Yale University.

— (forthcoming), A Reference Grammar of Bardi’, MS, Yale University.

Hercus, Luise, Flavia Hodges and Jane Simpson (eds) 2002, The Land is a Map: Placenames of Indigenous Origin in Australia, Pandanus Books, Canberra.

Keen, Ian 1995, Knowledge and Secrecy in Aboriginal Religion. Oxford University Press, Oxford.

Laves, G. (n.d.), Texts in Bardi, MS, held at Australian Institute of Aboriginal and Torres Strait Islander Studies, Canberra.

Metcalfe, Christopher Douglas 1975, Bardi Verb Morphology, Pacific Linguistics B30, Canberra.

— (n.d.), ‘Bardi Dictionary’, draft, handwritten dictionary of c.3000 items.

Robinson, Michael V. 1973, ‘Change and Adjustment amongst the Bardi’, MA Thesis, University of Western Australia, Perth.

Smith, Moya 1983, ‘Joules from Pools: Social and Techno-Economic Aspects of Bardi Stone Fish Traps’, in Archaeology at ANZAAS, Western Australian Museum, Perth: 29-80.

— 1984-5, ‘Bardi Relationships with the Sea’, Anthropological Forum, vol. 3: 443-447.

— 1987, ‘Dots on the Map: Sites and Seasonality, the Bardi Example’, Australian Archaeology, vol. 25: 40-52.

Worms, Ernst A. 1949, ‘An Australian Migratory Myth’, Primitive Man, vol. 22: 33-38.

— 1952, ‘Djamar and his Relation to other Culture Heroes’, Anthropos, vol. 47: 540-560.

— 1957, ‘Australian Mythological Terms: Their Etymology and Dispersion’, Anthropos, vol. 52: 732-768.

1 TResearch for this paper was carried out with funding from the Australian Institute of Aboriginal and Torres Strait Islander Studies (G2001/6505 and G2003/6761) and the Endangered Languages Foundation. Their support is gratefully acknowledged. I owe an immense debt to the Bardi people who have taught me their language and their placenames, who took me out to show me their country and who when we came back made me say where we had been to make sure I knew the placenames. An abbreviated version of this paper first appeared in Placenames Australia, June 2005.

2 The Native Title determination in June 2005 found that the first four areas under discussion (Goolarrgoon, Goowalgar, Baanarrad and Ardiyol) were traditionally Bardi, while Iwany and Mayala belonged to other groups; in particular, that Iwany belonged to Jawi people, who were found to be a distinct group at the time of European settlement in Western Australia (1829) and whose interests would not necessarily be served by the awarding of a claim where Bardi people would have primary jurisdiction over the land (since very few people these days identify as Jawi). I am not affiliated in any way with the land claim and strongly disagree with several aspects of the judgment, including the treatment of Bardi and Nimanburru as one cohesive social group on linguistic and cultural grounds, to the exclusion of Jawi. Therefore in this paper I am discussing both ‘Bardi’ country and ‘Jawi’ country, and treating Bardi and Jawi as two groups who shared ceremonies, language, and conventions for naming places.

3 The cognate word in Nyikina, ‘Wardiyabulu’, is the general word for the people belonging to the Northern Dampier Peninsula. The -yol portion of ‘Ardiyol’ is cognate with the -booloo of ‘Iinalabooloo’.

4 The name of this area in Robinson (1973: 103) is Baniol (i.e. Baniyol). I suspect that this represents a potential dialect difference between the Bardi people he worked with and those that I worked with.

5 It could be related to the word iwany ‘it is finished’ (i- third person singular present tense, -bany(i)- ‘finish, die’ (the lenition of b > w is regular in this position)), although this is contrived; there are no other placenames named after verbs. Earlier sources occasionally record the name as ‘Irrwany’.

6 The nasal is lost regularly before a stop. This word contains the same suffix (-bulu) as in Ardiyol.

7 Note also that jina booroo has grammaticalised in Yawuru to mean ‘someone’s niche, speciality’, like -iidi in Bardi. Bardi booroo thus has the semantic range of Arrernte apmere or Djirbal mija, with additional meanings (Hercus, Hodges and Simpson 2002). Booroo is also used interchangeably with the word baali ‘boughshed, shelter’, in the meaning of ‘camping place’.

8 The late Katie W. Drysdale (a senior Bardi woman born about 1905) used to make the comment that the name of this booroo used to be Manoogoon, not Ngamoogoon. I have been unable to verify this.

9 The elderly Bardi people I worked with no longer made regular visits to the country outside One Arm Point community, and when travelling along the road with them I was only ever given booroo names, not localities. In contrast, the late Nancy Isaac could recite hundreds of locality names from the coastal areas of her mother’s and father’s country, even from places she had not visited in more than 10 years. It may also be significant that the culture hero who names Bardi places in the text in (2) below goes ‘along the edge of everywhere’.

10 I have also recorded a variant, ‘Bulgu’, but it is not certain that this is the same place. ‘Bolg’ would be a spelling of the form which had gone through Nyulnyul sound changes, and the Bolg/Bulgu pair would be a counterexample to the claim that Bardi people use the name for the place from the language of the area.

11 This name is also spelt ‘Oogarrangg(oo)’ (the Bardi name for the people) and ‘Unggarrangg’.

12 The stories in particular involve the ‘Rainbow Serpent’ (or sea snake), whose name in Bardi is Oongoodoog. The word itself is not restricted and refers (for example, as in the lemma in Aklif 1999) to a large underwater serpent who lives off Sunday Island and occasionally steals people. Goodara’s mother, Wirrgoorr, was the last person to be taken in this way, in about 1910 – her story is told by Nancy Isaac in Bowern (ed.) (2002).

13 Ni- is the third person singular possessive prefix; the root is nga. The plural is irrnga.

14 The relevant maps are: 1:100 000 maps: 3465 Leveque, 3565 Sunday Island, 3464 Pender, 3564 Cornambie; 1:250 000 maps: SE51-2 Pender, SE51-3 Yampi. 1:1 million map: Broome SE51.

Previous Next