Temple sites in Kahikinui, Maui, Hawaiian Islands: their orientations decoded.

Kirch, Patrick V.

Introduction

With a few exceptions, mostly involving Easter Island (e.g. Ferdon 1961; Lee & Liller 1987; Liller 1989), archaeologists of Polynesia have largely ignored or avoided the subject of archaeoastronomy, a somewhat surprising situation considering that there is an abundance of ethnohistoric and ethnographic data regarding indigenous Polynesian and Oceanic knowledge systems bearing on astronomy, calendrics and navigation (e.g. Collocott 1922; Handy 1927; Henry 1928; Beaglehole & Beaglehole 1938; Makemson 1941; Hiroa 1932, 1938; Akerblom 1968; Valeri 1985). Over the past two decades, the experimental voyages of the double-hulled canoe Hokule'a, in particular, have brought renewed attention to Polynesian non-instrumental navigational abilities, which made extensive use of star knowledge (Finney 1994). In a recent synthesis of ancestral Polynesian culture, Kirch and Green (2001: 260-76) reconstruct key aspects of Polynesian time reckoning and the ancient ritual cycle. They conclude that all Polynesian cultures possessed:

* an annual seasonal cycle which divided the year into two parrs (Proto-Polynesian taqu), originally based on a wet-dry seasonality and the yam cultivation cycle;

* a sidereal cycle based on observations of the acronychal and heliacal rising of the star-cluster Pleiades (named Mataliki in Proto-Polynesian), which determined the transitions between seasons;

* an annual lunar (synodic) calendar of 13 months, which was keyed to the agricultural cycle; and

* a system of intercalation for keeping the lunar calendar synchronized with the solar year, based typically on the observation of Pleiades risings, but in some societies (especially Mangareva and Hawai'i) also on observation of solstice positions.

In Polynesian societies, astronomical and calendrical knowledge was the special purview of priests (and priest-navigators), who were responsible for making observations of key phenomena and for declaring when a particular season would end or commence. In Hawai'i, for example, the priests (kahuna) of the Lono cult closely awaited the first appearance of the Pleiades in the evening sky just after sunset towards the middle of November, which then marked the commencement of the Makahiki period of tribute collections and harvest celebrations (Handy & Handy 1972; Sahlins 1995; Valeri 1985). Similarly, the cult of the god Kane was closely associated with observation of the rising sun (Handy 1927; Handy & Handy 1972). Given that such priests were charged with regular observation of astronomical phenomena, it is not unreasonable to expect that the temples from which priests carried out their esoteric activities might also reflect this interest in astronomical phenomena, particularly in the positioning and orientation of temples.

Hawaiian temple orientations: prevailing perspectives

The pioneering archaeologist John F. G. Stokes, who carried out the first systematic and detailed architectural study of temples (heiau) on the islands of Hawai'i and Moloka'i in 1906-1909, wrote "I could find no evidence in the foundations of orientation to cardinal points. It is true that some of them did lie almost true north-south or east-west, but this was because the situation required it" (Stokes 1991:36). Rather, Stokes concluded that the orientation of heiau platforms was controlled by local topographic and environmental conditions (1991:35). Stokes' opinion was reaffirmed by later archaeologists. Wendell C. Bennett, who studied the heiau of Kaua'i, wrote that "true orientation to the points of the compass was seldom, if ever, considered, and was of little importance. The topography usually determined the orientation, the heiaus commonly facing the sea or valley" (1931:35; Emory 1924:62). For nearly a century, this has been the prevailing archaeological view. For only two cases, the isolated Ahu a 'Umi heiau on Hawai'i, and a group of heiau on Kaua'i, have proposals been advanced for orientations with astronomical significance (Da Silva & Johnson 1982; Ruggles 1999).

Hawaiian ethnohistory has a rich tradition deriving from a group of nineteenth-century Native Hawaiian scholars, whose writings provide a critical viewpoint on traditional Hawaiian practices. Unfortunately, these scholars are largely silent on the question of heiau orientations. Only David Malo, in his discussion of the construction of a luakini or war temple, offers a tantalising hint that heiau foundations may have been laid out according to cardinal directions: "The plan of the luakini was such that, if its front faced west or east, the lana-nuu-mamao [oracle tower] would be located at the northern end. If the heiau faced north or south, the lana-nuu-mamao would be located at the eastern part; thus putting the audience either in the southern or western part of the luakini" (Malo 1951:162). Malo's text hints that cardinal directions could have been important, and the positioning of the "audiences" in the south or west suggests that the main axes of orientation were therefore to the north and east. S. M. Kamakau (1976) provides a lengthy discussion of heiau construction, but says nothing of orientation per se. Nor do the shorter accounts of heiau building by Kamakau of Ka'awaloa and by S. N. Haleole, collected by Abraham Fornander (1919:56, 154-56), refer to orientation.

In sum, although archaeologists have observed empirically that heiau are at times oriented precisely north-south or east-west, in the absence of indigenous Hawaiian ethnohistoric texts that refer to preferred orientations, this has been taken to be coincidental. Instead, since Stokes' time it has been tacitly assumed that heiau foundations were laid out according to local topographical and environmental considerations, without regard for cardinal directions or astronomical phenomena.

The Heiau of Kahikinui, Maui

Since 1994 I have directed a long-term project focused on the ancient moku or district of Kahikinui, on the leeward slopes of Haleakala, Maui Island (Kirch ed. 1997). In collaboration with two other archaeological teams (Dixon et al. 2000; Kolb & Radewagen 1997), we have now acquired a systematic database that includes intensive survey data for two entire territorial units or ahupua'a (Kipapa and Naka'ohu), and from portions of four other ahupua'a (Luala'ilua, Alena, Naka'aha, and Mahamenui). Prior to our project, nine heiau sites had been reported for this area, largely through Winslow Walker's 1930 survey (Walker, ms.; Sterling 1998). For most of these no precise plans were available. We now have accumulated detailed survey records for some 30 heiau, many of which were previously unreported (see Kolb and Radewagen 1997 for an overview and for location maps).

Our survey methods include detailed plane table and alidade architectural mapping of the stone foundations of Kahikinui heiau (at scales of 1:50, 1:100, and 1:200, depending upon their size). Initially, I paid no particular attention to direction of orientation, although always recording N in the mapping procedure. By the 1998 field season, however, it was impossible to ignore the growing set of data indicating that Kahikinui temple foundations tended to have three dominant orientations: (1) towards the N, which is the direction of the 3,055-meter summit of Haleakala; (2) due E; and, (3) roughly E-N-E. This paper offers an analysis of our accumulated data on Kahikinui heiau orientations, and advances several hypotheses regarding their possible cultural significance. The analysis is based on a sample of 23 heiau for which we have accurate, large scale architectural maps, and can be confident with respect to orientation. These include sites in the ahupua'a of Alena (5), Kipapa (9), Naka'ohu (8), and Naka'aha (1). The sample includes a mix of coastal (6 sites), mid-elevation (5), and inland (12) sites. A list of heiau sites with key data on morphology, size, orientation, and viewsheds is provided in Table 1.

Chronology

Kolb and Radewagen (1997:76-77, table 5.3) reported seven radiocarbon dates for four heiau studied by them. With the exception of one anomolous early date from site HI-1386, most of their dates fall into the period from 1600-1800 cal AD. We have recently obtained 15 high-precision AMS radiocarbon dates from six heiau and an associated priest's residence, all clustered on a prominent ridge within Naka'ohu ahupua'a. These dates suggest initial construction of heiau in the sixteenth century AD, with others being constructed or used in the seventeenth-eighteenth centuries. Additionally, U/Th radiometric dating of branch coral used as offerings also confirms heiau use from the late sixteenth to mid-seventeenth centuries. In sum, all of the heiau within Kahikinui are likely to date within a three-century period from about 1500 to 1819 AD, the latter date being the date of the overthrow of the 'ai kapu or traditional religious prohibitions by the ruling chiefess Ka'ahumanu.

Size and morphological variation

As many have noted, Hawaiian temple foundations display enormous variation in plan and architectural details, and the Kahikinui heiau are no exception. Nonetheless, there is some overall patterning within this range of variation. In particular, we can identify three major morphological "types" of heiau in Kahikinui. The first is square in plan view, with enclosing stone walls, sometimes with a formal entryway. Typically the NE corner is wider and structurally elaborated, and there are often internal divisions. The second form is particularly common on Maui, and has been referred to as a "notched heiau" type (Figure 1), because in plan view it resembles a square from which a corner has been excised, resulting in a six-sided structure (Kolb 1994). Again, there is an enclosing stone wall, often more prominent (higher and/or thicker) on the E or N-E. Frequently there are various internal features, including pavements, altar terraces, and the like. The third category is here designated the "elongated two-court" heiau. These structures typically have a long central axis, with two distinct divisions or courts. One court is elevated above the other, and this elevated court is surrounded by a higher or more prominent wall than the lower court. The courts themselves are often slightly offset, again producing a kind of "notched" plan.

[FIGURE 1 OMITTED]

The sizes of Kahikinui heiau, measured in square meters of area, also show a wide range, as shown in Figure 2. There is one site which is much larger than any others, at 1400 [m.sup.2] (site 1010). The remaining sites display a spread in area from about 725 to only 55 [m.sup.2]. There is no clear-cut correlation between size and morphology; for example, "notched" heiau include the largest sire as well as some of the smallest. On the other hand, the "elongated, two-court" heiau tend to be among the largest temple foundations.

[FIGURE 2 OMITTED]

Landscape position and viewsheds

A further consideration when determining whether the Kahikinui heiau have preferred orientations is the local topographic setting, and the "viewsheds" from particular temple sites (Gaffney et al. 1996). Located on the leeward slopes of Haleakala, Kahikinui is not dissected by valleys, and the view planes from most of its prominent ridge lines offer sweeping perspectives across miles of open territory. Heiau were in many cases clearly situated to take advantage of these viewsheds. For example, the temples in Naka'ohu (the cluster we call "Heiau Ridge") are located on a prominent a'a lava flow ridge, with spectacular views towards Kaupo Gap and Kaupo Peninsula and to Kipahulu Valley beyond. These include many of the heiau with specific E or E-N-E orientations.

Orientation

With the exception of a few of the sites which are square in plan view, virtually all Kahikinui heiau display a dominant axis of orientation, indicated by one end of the structure being architecturally elaborated, with such features as raised altar platforms, thicker enclosing walls, elaborated buttressing, or multi-step terraces. This architecturally-elaborated end of the heiau is taken to be its direction of orientation. In cultural terms, this is the end of the heiau which is presumed to have been the sacred (kapu) direction, where images and offering platforms would have been located, and where priests would have officiated. The opposing end (180[degrees]) would have been the position of entry into the temple, and the seating location of what Malo (1951) called the "audience". This kind of axis is, in fact, typical of virtually all Polynesian ritual spaces, ranging from Hawaiian heiau to Rapanui ahu and Society Islands marae.

The direction of orientation was determined by measuring the bearing of such principal architectural features as the main enclosing walls, or terrace alignments. Bearings were taken with a magnetic compass (multiple readings were taken on all good wall faces), and corrected for the local declination of 11.5[degrees]E. Of course, Hawaiian dry stone masonry displays some variation, so that a "best fit" approximation of orientation direction usually had to be made. Although orientations are reported as a single reading, a margin of error of [+ or -] 2[degrees] is likely to obtain.

When the temple orientations are plotted on polar coordinates as a frequency distribution, by 10[degree] of arc, a striking, non-random pattern emerges (Figure 3a), with a strong preferred orientation towards the E, and with a subset tending E-N-E. Another preferred orientation is towards the N, or slightly N-N-W. When individual site orientations are plotted as in Figure 3b, it can be seen that there is: (1) a relatively tight cluster of nine sites with orientations between 82-93[degrees]; (2) a second cluster of five sites with orientations between 64-73[degrees]; and (3) a broader "spread" of eight sites with orientations ranging from 333-15[degrees]. Only one site falls outside of one of these clusters; this is site HI-482, a small notched heiau in Alena, with an orientation of 241[degrees]. It may be significant, however, that this small site is situated within a deep natural "ravine" in an a'a lava flow, with virtually no visibility of the surrounding landscape.

[FIGURE 3 OMITTED]

We have tested whether these orientation clusters or groups display positive correlations with either heiau morphology or size, with partly negative results. In terms of morphology, classic "notched" heiau are represented in all three orientation clusters, as are square heiau. However, the majority of "elongated" heiau are also those represented in the groups with an orientation of 333-15[degrees] (there is one exception, site A35-1, of an elongated heiau oriented virtually due E, 89[degrees]). Statistically, therefore, most of the elongated, two-court heiau face towards the high mountain of Haleakala. The results are similar when size is considered. The three orientation clusters are largely distributed across the size range of heiau; however, it is noteworthy that the three largest heiau are all with the 333-15[degrees] orientation group. Indeed, five of the seven largest heiau are within this roughly N-oriented cluster. However, lest one make too much of this correlation between size and orientation, it has to be pointed out that the smallest heiau in our sample is also in the 333-15[degrees] orientation group.

In sum, our empirical observations reveal the following: that the heiau of Kahikinui are not random with respect to orientation of their dominant architectural axes; that there are three orientation clusters, one facing the E, one centered at about 70[degrees], and a third oriented roughly N, which is towards the summit of Haleakala; and, that while there are no unambiguous correlations among orientation, size, and morphology, there is a statistical tendency for the N-oriented group to be both larger and of the "elongated", two-court style.

Interpretation: some working hypotheses

The empirical archaeological record reveals that Kahikinui heiau had definite and preferred axes of orientation. But, the question remains, how are we to interpret these given that the classic nineteenth-century Native Hawaiian sources are largely silent on the question of heiau orientations? These texts are not silent, however, with respect to traditional Native Hawaiian interest in cardinal directions, astronomical phenomena, or the reckoning of time. Samuel Kamakau devoted an entire chapter to a discussion of cardinal directions, the horizons, and the reckoning of time (Kamakau 1976:3-19). Of cardinal points, he writes that "it had always been customary with ka po'e kahiko, the people of old, to name the kukulu which are cardinal points in accordance with the rising of the sun in the east and its traveling to the west" (1976:3). His detailed discussion of these kukulu makes it clear that there was a complex and elaborate system of orientation with respect both to the ecliptic and to solar time, and to certain sidereal (star) positions. Malo offers a similar discussion of kukulu, writing that "the ancients named directions or points of the compass from the course of the sun" (1951:9). Kepelino (Beckwith, ed., 1932:80) adds additional corroboration:

 "The sun is the source by means of which the Hawaiians told
 direction on land because of the fixed boundaries it marks in the
 heavens. The place where the sun rises is the east, or hikina; the
 place where it sets is the west, or komohana. The place where Kiopaa
 stands is the north, or koolau; the side of the earth in the
 direction of the "cross-of-stars" is called the south, or kona. This
 is the old compass [ke panana kahiko] by which Hawaiians marked the
 position of places on land by the positions of the sun, the moon,
 and the important stars indicated." (Beckwith, ed., 1932:80).

In short, the pre-contact Hawaiians had a sophisticated system of cardinal directions, keyed to solar, lunar, and stellar phenomena. The same pertains to their system of temporal reckoning, which was a 13-month lunar calendar, annually corrected by sidereal and probably also solar observations (Valeri 1985:194-99; Kirch & Green 2001:267-73). This information provides a cultural context within which one may attempt to interpret Kahikinui heiau orientations.

If we also consider what is known of traditional Hawaiian theology, it is evident that the major deities of the Hawaiian pantheon were associated with particular directions and seasonal orientations. Table 2 reviews some of the associations between the four major deities of the Hawaiian pantheon and time, seasons, and other important phenomena, drawn from the extensive research of Valeri (1985:12-18, Table 1). Clearly, each deity had rather specific associations, including the following: Ku, the principal male god associated with war, canoe building, and sorcery, had for his directional attributes "high, east, and right," along with "high mountains" and "forest trees." Kane, who represents the male powers of procreation and was also the god of irrigated (taro) agriculture, had for his directions "right, east, and north". He was particularly associated with the rising sun, or as Kepelino put it, "the great sun of Kane" ["ka la nui o Kane"] (Beckwith, ed., 1932:81). Kane was likewise explicitly linked with the "Sun's northern limit on ellipse, summer." Lono, deity of dryland (especially sweet potato) agriculture, birth, and medicine, had directional attributes of "high, leeward", and was also linked to the annual rising of the Pleiades (Makali'i), the astronomical phenomenon which determined the onset of the annual harvest season, the Makahiki. Kanaloa, god of the sea and fishing, had directional attributes of "left, west, south" and was associated with the sunset and death.

This discussion, which could be elaborated further, provides the basis for a set of hypotheses regarding the possible significance of three orientation clusters identified for Kahikinui heiau. The notion that temples were functionally distinguished, with each temple being associated with a principal deity, is also well established in the ethnohistoric sources, (e.g. Hale a Lono or Unu o Lono temples associated with Lono, luakini or po'okanaka temples with Ku, ko'a shrines with Kanaloa, ho'oulu'ai with either Kane or Lono; see Valeri 1985:172-88).

The East cluster (82-93[degrees])

I propose that the nine temples which have E orientations were associated with Kane, the Hawaiian deity strongly associated with the sun and the E. The ecliptic or path of the sun across the sky was regarded as the path of Kane (ke alaula a Kane, "the flaming path of Kane"; Kamakau 1891; Makemson 1941:21). It is probable that from these heiau the priests of Kahikinui observed the progression of the sun throughout the year, and determined the architectural orientation of these temples by observation of the vernal and autumnal equinoxes. Viewed from the heiau cluster on the high ridgeline of Naka'ohu with its sweeping viewshed towards the east, the "great sun of Kane" would have risen dramatically over the peninsula of Kaupo.

The E-N-E cluster (64-73[degrees])

There are two astronomical phenomena which may have determined the orientation of the five temples in this cluster: the acronychal rising of the star-cluster Pleiades or Makali'i just after sunset, which determined the onset of the Makahiki season and the new year; and/or, the summer solstice. Azimuths for these phenomena have changed over time due to the precession of the equinoxes (Aveni 2001:100-103). Using a computer-generated sky chart as in Figure 4, however, we can determine that in the year 1750 AD, for example, Makali'i would first have been visible just after sunset (as opposed to rising later in the night) on about November 17, at an azimuth of 70[degrees]5'. In the same year, the sun would have risen on the summer solstice (June 22), at an azimuth of 67[degrees]4'. Both of these bearings are consistent with the orientations of this group of heiau.

[FIGURE 4 OMITTED]

Given the importance of dryland cultivation in Kahikinui, especially of 'uala or sweet potatoes (Ipomoea batatas), we can posit that Lono was an important deity to the people of this district. It seems probable that the heiau which are oriented ENE, at an azimuth of roughly 70[degrees], were Hale a Lono, or temples dedicated to this god. The Lono priests presumably awaited the rising of Pleiades in these temples, ready to announce the onset of the Makahiki and the new year. It is equally plausible that the summer solstice was also observed from these temples.

The North cluster (333-15[degrees])

Finally, what of the N-oriented cluster of heiau? This group of eight heiau displays a greater spread of azimuths than the other two clusters, and we need to keep in mind that to be oriented roughly N in Kahikinui is to face the great soaring height (3055 m high) of Haleakala ("House of the Sun"). Most likely, the temples were constructed so that their principal axes faced the mountain, either its high summit or one or other of the prominent, reddish-colored cinder cones (pu'u) that stand out along the skyline of the mountain's south-western rift zone.

A majority of the "elongated, two-court" temples, as well as many of the largest temples, have this northern orientation. The hypothesis favored here is that these temples were dedicated to, or associated with, Ku, the deity linked to the high mountains, to the sky, and to forests. Ku is often thought of primarily as the god of warfare, and indeed he was certainly the main deity of the ruling chiefs with their political aspirations. But in the marginal, "back country" of Kahikinui, it is conceivable that the worship of Ku was focused not so much on his warlike propensities, but on some of his other attributes which included canoe building and sorcery.

Conclusions

Contrary to the orthodox view of several generations of archaeologists who have held that Hawaiian temples were laid out to best fit the local land contours and were not purposively oriented to cardinal directions or particular astronomical phenomena, the known sample of heiau in Kahikinui clearly have preferred orientations, including E, E-N-E, and N. I have offered a set of working hypotheses which associates each of these preferred orientations with a particular set of astronomical or landscape phenomena, and with particular deities of the traditional Hawaiian pantheon. It would be premature to leap to the conclusion that all temples throughout the archipelago were also oriented preferentially; this will require further empirical study. For example, recent archaeological survey in the Kalaupapa region of Moloka'i Island has also revealed the presence of several temple sites with E and E-N-E orientations (Kirch, ed., 2002).

It remains likely that many heiau were oriented according to landscape features and to take advantage of particular viewsheds, rather than to cardinal directions. Nonetheless, the Kahikinui evidence suggests that it will be productive to revisit the issue of heiau orientations more widely in the Hawaiian Islands. Moreover, this study raises the question of whether-in light of the considerable body of ethnohistoric and ethnographic data on Polynesian astronomical, calendrical, and navigational systems of knowledge-archaeologists may wish to pay more attention to the possibilities that ritual sites throughout the Polynesian archipelagoes may frequently have been preferentially oriented.

Table 1 Data on Kahikinui Heiau

 Axis of
Site Size Orientation
Number * Ahupua'a Morphology ([m.sup.2]) ([degrees]E)

HI-1157 Luala'ilua Elongated, 2-court 720 350

HI-183 Luala'ilua Square 210 351
 (internal divisions)

HI-184 Luala'ilua Notched 59 340

HI-482 Alena Notched 79 241

HI-178 Alena Elongated, 2-court 472 333

A35-1151 Alena- Square, terraced 182 82
 Kipapa
 boundary

A35-1 Kipapa Elongated, 2-court 625 89

A35-728 Kipapa Elongated, 2-court 704 4

A35-1307 Kipapa Notched variant 88 87

A35-1156 Kipapa Square 210 73

A35-1146 Kipapa Terraced, irregular 300 92

A35-307 Kipapa Notched 460 345

A35-273a Kipapa Notched 92 93

A35-273b Kipapa Rectangular, terrace 82 83

A35-414 Nakaohu Notched variant 433 92

A35-80 Nakaohu Terraced platform 200 15

A35-405 Nakaohu Complex form 400 88

A35-115 Nakaohu Complex form 500 70

A35-77 Nakaohu Notched 327 71

A35-75 Nakaohu Square 225 69

A35-567 Nakaohu Square 126 64

A35-1025 Nakaohu Notched variant 500 87

A35-1010 Nakaaha Double notched 1400 0

Site
Number * Viewshed

HI-1157 Views to Haleakala and to coast

HI-183 Views to Haleakala and to coast

HI-184 On coast, view to Haleakala

HI-482 In lava depression, no views

HI-178 Limited viewshed to coast

A35-1151 Lava cone to E, views of Alena
 to W

A35-1 Low rise to E, views of Haleakala

A35-728 View of Haleakala, portions of
 Kipapa

A35-1307 Limited view over agricultural
 swale

A35-1156 Overlooks stream channel, views
 to Haleakala and portions of
 Kipapa

A35-1146 Views over lower Kipapato coast,
 Haleakala

A35-307 On coast, views E and W along
 coast and to Haleakala

A35-273a View to Haleakala

A35-273b Coastal view to W

A35-414 Broad E view, Haleakala

A35-80 Sweeping view over Kipapa
 ahupua'a

A35-405 On prominent ridge, broad E
 view to Kaupo, Haleakala visible

A35-115 On lava ridge, view to E over
 Nakaohu, Haleakala visible

A35-77 E view over agricultural zone,
 Haleakala visible

A35-75 View to Haleakala

A35-567 E view over Nakaohu, Haleakala
 visible

A35-1025

A35-1010 Sweeping view E to Kaupo and of
 Hawaii Island, Haleakala visible

* Numbers with the prefix HI--are in the State of Hawaii site inventory
system, while those with a A35--prefix are in the Bishop Museum system.

Table 2 Some Attributes of the Major Hawaiian Deities *

Attribute Ku Kane

Directions East, high, right Fast, north, right

Calendar First 3 days of lunar Dawn, 27-28th days
 month, dawn of lunar month

Landscape & High mountains, Light, sun, fresh
natural phenomena sky waters

Seasons Season of kapu pule Summer, sun's
 (temple ritual) northern limit
 on ellipse

Functions War, canoe building, Male power of
 sorcery procreation,
 irrigated agriculture

Attribute Lono Kanaloa

Directions High, leeward South, west, left

Calendar 28th day of lunar 23-24th days of lunar
 month month, sunset

Landscape & Thunder, rain Seawater, ocean
natural phenomena

Seasons Rising of Pleiades, Winter
 Makahiki season

Functions Dryland agriculture, Fishing, death
 medicine

* Data abstracted from Valeri (1985, table 1).

Received: 6 December 2002 Accepted: 27 June 2003 Revised: 11 August 2003

References

AKERBLOM, K. 1968. Astronomy and Navigation in Polynesia and Micronesia. Etnografiska Museet, Publication 14. Stockholm: The Ethnographical Museum.

AVENI, A. F. 2001. Skywatchers. University of Texas Press. Austin.

BEAGLEHOLE, E. & P. BEAGLEHOLE. 1938. Ethnology of Pukapuka. Bernice P. Bishop Museum Bulletin 150. Honolulu.

BECKWITH, M. W. (ed). 1932. Kepelino's Traditions of Hawaii. Bernice P. Bishop Museum Bulletin 95. Honolulu.

BENNETT, W. C. 1931. Archaeology of Kauai. Bernice P. Bishop Museum Bulletin 80. Honolulu.

COLLOCOTT, E. E. V. 1922. Tongan astronomy and calendar. Occasional Papers of the B. P. Bishop Museum 8:157-73.

DA SILVA, A. M. & R. K. JOHNSON. 1982. Ahu a 'Umi Heiau: A Native Hawaiian astronomical and directional register. A. F. Aveni & G. Urton, eds., Ethnoastronomy and Archaeoastronomy in the American Tropics, pp. 313-31. Annals of the New York Academy of Sciences 385. New York.

DIXON, B., P. J. CONTE, V. NAGAHARA & W. K. HODGINS. 2000. Kahikinui Mauka: Archaeological Research in the Lowland Dry Forest of Leeward East Maui. Report prepared for the Department of Hawaiian Home Lands. Kahului: Historic Preservation Division, State of Hawaii.

EMORY, K. P. 1924. The Island of Lanai: A Survey of Native Culture. Bernice E Bishop Museum Bulletin 12. Honolulu.

FERDON, E. N., Jr. 1961. The ceremonial site of Orongo. T. Heyerdahl & E. N. Ferdon, Jr., eds., Archaeology of Easter Island. 221-55. Monographs of the School of American Research and the Museum of New Mexico, Number 24, Part 1. Chicago: Rand McNally & Co.

FINNEY, B. 1994. Voyage of Rediscovery: A Cultural Odyssey through Polynesia. Berkeley: University of California Press.

FORNANDER, A. 1919. Fornander Collection of Hawaiian Antiquities and Folk-lore. (T G. Thrum, ed.) Memoirs of the Bernice E Bishop Museum, Volume VI. Honolulu.

GAFFNEY, V., Z. STANCIC & H. WATSON. 1996. Moving from catchments to cognition: Tentative steps toward a larger archaeological context for GIS. M. Aldenderfer & H. D. G. Maschner, eds., Anthropology, Space, and Geographic Information Systems: 132-54. New York: Oxford University Press.

HENRY, T. 1928. Ancient Tahiti. Compiled from Notes of J. M. Orsmond. Bernice P. Bishop Museum Bulletin 48. Honolulu.

HANDY, E. S. C. 1927. Polynesian Religion. Bernice P. Bishop Museum Bulletin 34. Honolulu.

HANDY, E. S. C. & E. G. HANDY. 1972. Native Planters in Old Hawai'i: Their Life, Lore, and Environment. Bernice P. Bishop Museum Bulletin 233. Honolulu.

HIROA, T. R. (P. H. BUCK). 1932. Ethnology of Manihiki and Rakahanga. Bernice E Bishop Museum Bulletin 99. Honolulu.

--1938. Ethnology of Mangareva. Bernice P. Bishop Museum Bulletin 157. Honolulu.

KAMAKAU, S. M. 1891. Instructions in ancient Hawaiian astronomy as taught by Kaneakahoowaha. Thrum's Hawaiian Annual for 1891, pp. 142-43. Honolulu.

--1976. The Works of the People of Old: Na Hana a ka Po'e Kahiko. Bernice P. Bishop Museum Special Publication 61. Honolulu.

KIRCH, P. V. (ed.) 1997. Na Mea Kahiko o Kahikinui: Studies in the Archaeology of Kahikinui, Maui. Berkeley: Archaeological Research Facility, University of California.

--2002. From the "Cliffs of Keolewa' to the "Sea of Papaloa': An Archaeological Reconnaissance of Portions of the Kalaupapa National Historical Park, Moloka'i, Hawaiian Islands. Berkeley: Archaeological Research Facility, University of California.

KIRCH, P. V. & R. C. GREEN. 2001. Hawaiki, Ancestral Polynesia: An Essay in Historical Anthropology. Cambridge: Cambridge University Press.

LEE, G. & W. LILLER. 1987. Easter Island's "sun stones": A re-evaluation. Archaeoastronomy 11:S1-S11.

LILLER, W. 1989. The megalithic astronomy of Easter Island: Orientations of ahu and moai. Archaeoastronomy 13:S21-S48.

KOLB, M. J. 1994. Monumentality and the rise of religious authority in precontact Hawai'i. Current Anthropology 35:521-47.

KOLB, M. J. & E. C. RADEWAGEN, 1997. Na Heiau o Kahikinui: The Temples of Kahikinui. P.V. Kirch (ed.) Na Mea Kahiko o Kahikinui: Studies in the Archaeology of Kahikinui, Maui: 61-77. Berkeley: Archaeological Research Facility, University of California.

MALO, D. 1951. Hawaiian Antiquities (Moolelo Hawaii). Bernice E Bishop Museum Special Publication 2. Honolulu.

MAKEMSON, M. W. 1941. The Morning Star Rises: An Account of Polynesian Astronomy. New Haven: Yale University Press.

RUGGLES, C. 1999. Astronomy, oral literature, and landscape in ancient Hawai'i. Archaeoastronomy: The Journal of Astronomy in Culture 14:33-86.

--2000. Ancient astronomies-ancient worlds. Archaeoastronomy 25:65-76.

SAHLINS, M. 1995. How "Natives" Think: About Captain Cook, For Example. Chicago: University of Chicago Press.

STERLING, E. P. 1998. Sites of Maui. Honolulu: Bishop Museum Press.

STOKES, J. F. G. 1991. Heiau of the Island of Hawai'i: A Historic Survey of Native Hawaiian Temple Sites (T. Dye, ed.). Bernice P. Bishop Museum Bulletin in Anthropology 2. Honolulu.

VALERI, V. 1985. Kingship and Sacrifice: Ritual and Society in Ancient Hawaii. Chicago: University of Chicago Press.

WALKER, W. M., ms. [1930]. Archaeology of Maui. Typescript in Archives, Bernice P. Bishop Museum, Honolulu.

Patrick V. Kirch (1)

(1) Department of Anthropology, and Archaeological Research Facility, University of California, Berkeley CA94720, USA. (Email: kirch@sscl.berkeley.edu)