Volume 80 1971 > Volume 80, No. 4 > Radiocarbon dates from Easter Island, by William S. Ayres, p 497 - 504
RADIOCARBON DATES FROM EASTER ISLAND
Easter Island, the easternmost extension of Polynesia, at 109° 25′ west longitude and 27° 08′ south latitude, has long been of archaeological interest primarily because of the impressive stone monuments called ahu concentrated around the coast of the island. The structures called ahu on Easter are the structural and functional equivalent of the marae on other Polynesian islands. The first major attempt to examine systematically the archaeological remains of the island came with the Heyerdahl expedition in 1956. Results of excavations in a variety of sites and 18 radiocarbon dates from archaeological contexts were published. 1 More recent field research by William Mulloy has added to our knowledge of the structure and chronology of ahu. 2
Six additional radiocarbon dates are now available to assist in outlining the prehistory of Easter Island. The samples submitted to Gakushuin University were collected by the author during field work on Easter Island in 1968 under the direction of William Mulloy. 3
Field research was conducted from March to August, 1968, at two ahu in a locality known as Tahai, on the west coast of the island, about one kilometre north of the present village of Hangaroa. Three ahu forming a structural unit within a larger ceremonial complex are Ahu Ko te Riku, also known as Ko te Ahu o te Vaka Ariki (Site 8-1), Ahu Tahai (Site 8-2), or Ko te Ahu o te Nga'ara Ariki, and Ahu Vai Uri (Site 8-3). The two ahu discussed here, Ko te Riku and Tahai, had only one statue each. There are structural indications that these three ahu, forming a unit covering over 20,000 square metres around a small bay, were used contemporaneously during at least part of the period Smith has defined as the Middle Period of Easter Island prehistory. 4 This period was characterised primarily by the presence of large statues carved of tuff placed on the central platform of the ahu. Smith associated dates of ca. A.D. 1100 to 1680 with this period.- 498
In general, the six new dates seem to support the broad chronology, based primarily on structural changes in ceremonial centres of the ahu type, set up by Smith as a result of the Heyerdahl expedition's research. 5 Of the following dates, four were obtained for Ko te Riku and two for Ahu Tahai.
An analysis of the true age and significance of these dates in relation to their context is necessary.
Sample number 3 (GaK-2862) from Ko te Riku was taken from a charcoal lens within the fill of the north wing of the ahu and dates the construction of the north wing, and by extension, the entire ahu, since excavations revealed that the structure was built as one unit. The sample was found 22 to 27 cm below the ground surface at one edge of a section of undisturbed paving stones covering the wing fill of stone and soil. The radiocarbon age (910 ± 90 B.P.) for this sample was earlier than had been expected, since the nature of construction pointed to a late date within the time span indicated by previous field work for similar ahu of the Middle Period such as Vinapu 1, 440 ± 100 B.P. (thought to date construction of the Middle Period structure), and Ahu Akivi, 425 ± 100 B.P. 6 The Middle Period type ahu, in generalised form, has been thoroughly defined in the reports of the Heyerdahl expedition; 7 thus only a brief description of the type will be needed here. This ahu consists of a rectangular-shaped central platform constructed of both cut and uncut stone. Cut stone taken from earlier structures, where it was available, was often re-used in the ahu of this later Middle Period without refitting. Lateral extensions, or “wings”, on either side of this platform, slope down towards the court or plaza level on the landward side. The inclined ramp of fill in front of the central platform and wings was paved with rows of water-worn stones. The most distinctive feature of the Middle Period ahu, as it has been defined at this stage of our research in Easter Island prehistory, is the presence of large statues of volcanic tuff on the central platform.
Conversion of radiocarbon age from radiocarbon years before present (1950) to calendrical years is no longer thought to be a simple one-to-one ratio because of the long-term variation in the C14 level. The true ages of the samples given here are corrected according to the determinations of Stuiver and Suess. 8 An additional complication in conversion, noted by Stuiver and Suess, is that a radiocarbon age may correspond to more than one true age. 9 A minimum standard deviation is taken as 100 following the suggestion of Polach and - 499 Golson. 10 In the case of sample number 3 (910 ± 90 B.P.) the possible true ages are 745, 770 or 840, ± 90 B.P. according to the graph of Stuiver and Suess. 11 Since there are no other indicators at present to allow choosing one of these values as the correct mean, in this case it seems best to refer to a time range that covers all the expressed possibilities. This would be from 745 to 840 B.P. or A.D. 1110 to 1205 ± 90. The lower limit at one standard deviation, taken as a minimum of 100, from 840 B.P., gives 940 B.P. or A.D. 1010. The upper limit at one standard deviation would be 645 B.P., which is A.D. 1305. Thus the time of construction of the ahu can only be put within the time period of A.D. 1010 to 1305 for one standard deviation.
The date obtained for Ahu Akivi of 425 ± 100 B.P. would be corrected to A.D. 1455 ± 100 (there is only one true age for this radiocarbon age) and for Vinapu 1, a date of 440 ± 100 B.P. would be corrected to A.D. 1450 ± 100 for a similar structure of the Middle Period (see Figure 1). These two dates are those of construction of the Middle Period structures at the two ahu. 12
Sample number 5 from Ko te Riku (GaK-2863) is a charcoal sample from the bottom of a firepit dug into the north wing of the ahu. An initial determination by the Laboratory of 1890 ± B.P. for this sample was considered to be erroneous, because of a presumed mix-up in processing. The second determination of 880 ± 70 B.P. is still earlier than expected. It should have dated considerably later than sample number 3 discussed above, which dates construction of the wing into which the firepit was dug. The pit contained three large cut stones (one a foundation stone of a boat-shaped house) and a heavy concentration of charcoal and ash. The slight depression left by the filled in pit covered about nine square metres in the centre of the wing (three by three metres). The true age of the activity should post-date the 910 ± 90 B.P. result corrected to A.D. 1010 to 1305, since it was probably a fairly recent earth oven made by digging a hole into the fill of the north wing of the ahu after it was no longer in use. The corrected age for radiocarbon ages this old, according to Stuiver and Suess, is 810 or 740 ± 70 B.P., or a range of A.D. 1040-1310 at one standard deviation, 13 but this is still not considered young enough to be accurate. The bottom of this firepit was near the original ground surface beneath the fill at a depth of 45 to 52 cm below the top of the wing fill. The sides of the pit were clearly visible in the profile.
Sample number 9 (GaK-2864), from Ko te Riku, was taken from a charcoal lens beneath a dome-shaped, paved mound at the north edge of the ahu. The date of 1010 ± 90 B.P. is from a concentration of charcoal on a rock outcrop with associated hammerstones for flaking volcanic glass and considerable detritus from this activity. The corrected age is 50 years younger than the uncorrected A.D. 940 ± 90 age indicated by the radiocarbon age determination. 14 This can be expressed as 960 ± 90 B.P., or a range of A.D. 890-1090 at one standard deviation. The sample appears to pre-date construction of the mound by only a short time since it is just below the ground surface upon which the stone and soil fill of the mound was placed. The date range is close enough to that of construction of the ahu to indicate that the mound was probably constructed at the same time. The function of the mound remains obscure.
Sample number 16 is a sample of cremated bones from a cremation pit on the plaza of Ko te Riku. The result of 780 ± 90 B.P. (GaK-2865), corrected to - 500 A.D. 1210 ± 90, is similar to other dates for cremation that have been reported for Easter (see Figure 1). 15 The pavement of the plaza had been disturbed so it is not possible to tell definitely if the cremated bone deposits pre-date the pavement. The time range given for construction of the ahu, A.D. 1110 to 1205 ± 90, compared with the date for the cremation sample 16 (A.D. 1210 ± 90), suggests that they were contemporaneous with use of the ahu, and therefore post-date the paving. The calvarium of a young female adult had been interred in a small stone cist at one edge of the cremation area. This could account for some of the pavement disturbance. A mataa, considered mainly a Late Period artefact (ca. A.D. 1680-1868), 16 was associated with the skull. The sample from the cremation pit of Ahu Tahai gave an identical date and supports the contention that the two ahu were used simultaneously.
Only two radiocarbon samples were submitted for Ahu Tahai. Sample number 1 was of charcoal and burned earth from within the fill of the first stage of the ahu. This stage will be designated Tahai I. The sample was found 25 to 35 cm below the surface of the level paving that extended in front of the central platform and over the wings of the first period structure, and so can definitely be associated with construction of Tahai I. This sample location is about three metres south of the central platform in the fill of the wing which extended out horizontally from the central platform. Depth below the surface at the time of excavation was 35 to 45 cm. Construction of a windmill over the south wing of the ahu in the 1940s had levelled the fill of the superimposed second phase, Tahai II, to within 10 cm of the pavement below which the sample was taken. This sample yielded a date of 1260 ± 130 B.P. (GaK-2866) or A.D. 690 ± 130. Stuiver and Suess indicated that seventh century first millenium dates may not need the 50- to 100-year correction factor which is considered necessary for calendrical ages of other centuries of that millenium. 17 This is, then, the earliest reported date for an ahu of the Early Period as defined by Smith. 18 The features of this stage for Ahu Tahai are similar to those described for other known early structures such as those underlying later constructions at Vinapu 1 and 2 and Ahu Akivi. 19 Tahai I, however, appears to have been started much earlier than Ahu Akivi or Vinapu I and perhaps earlier than Vinapu 2.
Sample number 6 from Ahu Tahai is a sample of charred bone from a crematorium or depository for cremated bones in the surface of the paved ramp in front of (landward side) the central platform of the second stage of Ahu Tahai, Tahai II. This second stage is a Middle Period structure as shown by its form and the presence of a large tuff statue placed on the central platform. The pit, about 25 cm deep and 58 cm wide by 70 cm long, is enclosed by a border of cut slabs of red scoria. The radiocarbon age of 810 ± 80 B.P. has been corrected to give a calendrical date of A.D. 1210 ± 80. This date, as previously mentioned, is the same as that for the cremation pit on the plaza of Ko te Riku. However, these bone dates should be accepted with caution since use of the cremation pit over a fairly long period may have caused mixing of the deposits. The general unreliability of carbon-content bone dating itself may lead to error. A small iron nail found at a depth of 20 to 40 cm below the surface in the rock fill of the north wing of the Ahu Tahai II structure may indicate that the cremation date is too early, but it is possible that the nail could have fallen through between the rocks after construction. If the nail is taken to indicate that the structure- 501 - 502
was built in historical times, the date for the cremation activity would be several hundred years too early, which seems unlikely considering the early dates for the underlying Tahai I structure and for Ahu Ko te Riku. A date range of A.D. 1110-1310 at one standard deviation is therefore indicated for this sample. The second stage of the ahu, Tahai II, was built some time before the period of the cremations.
We are now in a position to examine the significance of the corrected dates. It appears that of the two ahu, Tahai and Ko te Riku, Tahai I was the first structure built and that activity could have taken place several hundred years before construction of Ko te Riku and even of the superimposed Tahai II structure. The time range for Tahai I is A.D. 560 to 820 at one standard deviation; that is, there is a fair probability that it was built between these dates. Radiocarbon dates are not yet available for the third ahu of the complex, Vai Uri. The next building activity at this site was probably the construction of Tahai II over Tahai I, which was in a state of disrepair at that time: some time before A.D. 1310 and perhaps before A.D. 1100. The range of A.D. 1100-1310 is that given at one standard deviation for the cremation pit dug into the surface fill of Tahai II which must post-date construction of that structure. Lines of paving stones disturbed in placing the border of the cremation pit indicate that it was not built at the same time as Tahai II itself, and could be considerably later.
Ahu Ko te Riku was also built during the same period, between A.D. 1010 and 1305. The period of construction of this ahu probably also saw the building of the paved mound at the north end of Ahu Ko te Riku, covering an earlier firepit (dated between A.D. 890 and 1090). Evidence that the cremation pit on the plaza of Ko te Riku was used between A.D. 1110 and 1310 indicates that it was probably contemporaneous with the ahu.
Figure 1 illustrates the relation of the corrected dates associated with ahu structures. Dates from seven different ahu are included in this graph. These are all the published dates from ahu structures except for the one of 1640 ± 250 B.P. (M-732) obtained by Smith from Ahu Tepeu, which he considers to be in error. 20 The corrected mean dates are given with both one and two standard deviations indicated. In cases where more than one mean value is obtainable from the Stuiver and Suess graph or tables, the upper and lower mean values are used as a time range to cover the inclusive possible ages. This is represented as an open bar on the graph.
The sequence of the two sites strengthens the validity of the chronology proposed by Smith as a result of the Heyerdahl expedition's research. The proposed settlement date of ca. A.D. 400 still seems reasonable in light of this new information since Tahai I was a large, complex structure at the early date shown by radiocarbon analysis. An organised work force of considerable size would have been necessary. This date for settlement also seems entirely compatible with present understanding of the relationships of Polynesian languages. 21 There appears to be a valid distinction between an Early Period extending to about A.D. 1100, with ahu structures emphasising cut stone and the absence of statues on the central platform, and a Middle Period characterised by an emphasis toward raising increasingly larger statues, all carved at the same quarry, on to the central platform. This distinction, however, appears to be the reflection of a trend rather than a discontinuity of culture. The present excavations give no clear evidence of statues associated with Tahai I, but do indicate considerable cut stone. However, the head of a small statue of red scoria, carved in a naturalistic style, found in the bay at the seaside of the ahu, is - 503 believed to have been associated with Tahai I. The Tahai I structure does appear to have been oriented significantly with the winter solstice. Solar orientation has been proposed for Early Period ahu by both Mulloy and Smith. 22 The overall continuity of ahu style and the presence of small statues associated with earlier structures at several locations lead the author to believe that Easter Island ahu will eventually be shown to have developed from a small platform with small statues located either on the plaza or on the platform itself, to ahu with rapidly increasing emphasis on the size of the statue placed on the elevated central platform. Considering the very few ahu that have so far been thoroughly examined, however, any chronology based on them is quite speculative.
One of the pressing needs is to determine whether the developmental continuity proposed above can also be found in artefact types.
A more detailed report on the field work discussed above is in preparation.
Six radiocarbon samples from Easter Island, East Polynesia, were analysed by Gakushuin University, Japan, with funds from the Wenner-Gren Foundation. The samples, collected by William S. Ayres during field work under the direction of William Mulloy in 1968, were selected to give dates for construction and use of the two ceremonial centres, or ahu, that were excavated and restored. These dates support the broad chronology proposed by Carlyle Smith of the Heyerdahl expedition for ceremonial structures of this type on Easter Island. The time range for use of the site as a ceremonial centre was from A.D. 690 ± 130 to the late 1800s. All of the radiocarbon ages were corrected using the calculations of Stuiver and Suess to give calendrical dates. The first ahu constructed was Ahu Tahai (Tahai I). Later, during the statue-bearing phase of ahu use, a rebuilding of Tahai (Tahai II) was joined with the construction of Ahu Ko te Riku and also with the third structure of the unit, Ahu Vai Uri. The radiocarbon dates support this view. Analysis of these dates indicates that the proposed date of ca. A.D. 400 for the initial settlement of the island seems reasonable. The sharp distinction made previously between the Early Period and the Middle Period ahu is less clear now on the basis of increasing archaeological knowledge.
1 Smith 1961b:393-6.
2 Mulloy and Figueroa 1962; Crane and Griffin 1965:146-7.
3 The author wishes to acknowledge the great assistance and the opportunity to do research on Easter Island given by William Mulloy, Professor of Anthropology, University of Wyoming. Mulloy was director of an archaeological venture, of which the author was a member, on Easter Island, sponsored jointly by the International Fund for Monuments of New York, N.Y., and the Chilean Government. An expression of appreciation is also due to Arden R. King, Professor of Anthropology, Tulane University, for continual assistance and for aid in obtaining funds for processing the samples. The funds for dating were received from the Wenner Gren Foundation as an aid to dissertation research.
4 Smith 1961a:210-13.
5 Smith 1961a.
6 Mulloy 1961:99, 160; Crane and Griffin 1965:146.
7 Smith 1961a; Ferdon 1961:527-35.
8 Stuiver and Suess 1966:537-8, Table 1, Fig. 1; see Green 1970 for an earlier use of their corrections.
9 Stuiver and Suess 1966:537.
10 Polach and Golson 1966:22.
11 Stuiver and Suess 1966: Fig. 1.
12 Mulloy 1961:160; Crane and Griffin 1965:146.
13 Stuiver and Suess 1966:537, Fig. 1.
14 Stuiver and Suess 1966:537, Fig. 1.
15 Mulloy 1961:100, 160; Crane and Griffin 1965:146.
16 Ferdon 1961:531.
17 Stuiver and Suess 1966:536.
18 Smith 1961a.
19 Mulloy 1961:100-05, 119-21, 160, and personal communication.
20 Smith 1961b.
21 Elbert 1953; Green 1966.
22 Mulloy 1961:159; Smith 1961a:218-19.