Before continuing from where I left in the previous input, I should probably insert a little explanatary note for those members who are not immediately familiar with some archaeological terms. On rereading (1a) I noted I used the term "neolithic" in a way which might not be understandable by someone who was not informed that it is the last or upper period of the stone age, being in contrast with the earlier paleolithic period a.o. in that there was agriculture or horticulure. To express it popularly: the neolithic period is the part of the stone age where there already is agriculture. That said, I can now continue.
The next stage is a bit problematic, because there are two opinions as to the beginnings of the planting of cereals by Austronesians. The formerly established opinion was, that Austronesians originally planted tubers as starch staple, particularly taro (Colocasia esculenta; Indonesian talas = keladi), yams (Dioscorea spp.; Indon. ubi), alocasia (Alocasia microrhiza and other spp.; Indon. birah). Only at the later period did they start planting rice (Oryza sativa; Indon. padi), explaining why the distribution only goes as far east as to include East Indonesia, but not Oceania. If one stopped here, one could conclude another important difference in the culture of the two regions, Maluku and part of Irian Jaya (North, up to Cendrawasih Bay inclusively) with its at least partially cereal agricultural economy, and Oceania, particularly (the rest of) Melanesia with an agriculture economy based on cultivation of tubers.
However, that old picture of the development does not account for some particularities uncovered by closer linguistic study and newer archaeological discoveries. Reflexes of the protoform *pajei (hence also Malay padi) are not only well represented in languages of Indonesia and the Philippines, but also in those of Taiwan. The protoforms of the names of the tubers (*tales, *quBi, *BiRaq), on the other hand, being well represented throughout Oceania and Indonesia, partly also the Phlippines, do not have reflexes in languages of Taiwan. Insofar as that the highest order (earliest) split in the Austronesian language family was between that of the Formosan languages (Austronesian languages of Taiwan) from Extra-Formosan Austronesian languages, a protoform represented in Formosan as well as Extra-Formosan languages must originate from Proto-Austronesian. In other words, the Proto-Austronesians must have been acquainted with rice, and were probably not acquainted with the tubers or at least not intensively enough for these to have left a lasting imprint in the language.
From biological data its is known, furthermore, that wild varieties of the tubers in question are endemic in regions in mainland and insular Southeast Asia. Rice, on the other hand, was first domesticated in a long zone at the northern rim of Indochina and Assam, and southern rim of presentday China. Archaeological data show that it was already cultivated in Zhejiang (the mainland province across from Taiwan) by 5000 BC (Peer Bellwood). This is well before the date of first split of Proto-Austronesian, estimated at 4000 BC (Robert Blust) by linguists, and slightly later (Peter Bellwood) by archaeologists. On Taiwan itself, however, earliest trace of rice in the form of imprints on pottery is dated at around 3000 BC (Peter Bellwood). If indeed rice had been cultivated by the Proto-Austronesians intensively enough to bo leave such a clear trace in the language, one needs to find an explanation for the absence of cereal cultivation in pre-contact Oceania.
There is one important feature in rice cultivation, which is probably crucial for the understanding of the particularities of its dispersal through Austronesian migrationary and contact movements. For its ripening, original varieties of rice required a longer daily photoperiod than that which is given in the tropical zone (the day in non-tropical zones is longer than in the tropics in summer, when rice ripens, and shorter in winter). Therefore, the original varieties of rice were not suitable for the tropics. Even if the Austronesians had been rice-growers since the very beginning, arriving in the Philippines (where they are already reported in the Cagayan valley of Luzon around 3000 BC, Robert Blust) they must have been forced to look around for a new starch staple before moving on. Indeed, they appear not only to have taken up the already mentioned tubers here, but also the coconut (Cocos nucifera; reflexes of *nieur, e.g. Indon. nyiur, well represented all over Oceania and insular Southeast Asia, are missing in Taiwan). In other words, it must have been only after a total agricultural re-equipment that they proceeded onwards into the tropics and further into Oceania. This will perhaps also help us understand, why it was that Hendrik Kern, determining the homeland of the Austronesians on the basis of their common zoological and botanical vocabulary, decided over a century ago that this must have been in tropical Southeast Asia.
In a 1994 publication I suggested on the basis of various considerations, that the first rice to have been spread through the Philippines and Indonesia was a highland variety, and that even then it was not cultivated as main staple. It was just sufficiently present to have its name propagated mostly in phonologically regular form (i.e. not through later borrowing which would have rendered the sound correspondences irregular) throughout the Philippines and Indonesia, without it having a role of any importance in the economy. The first cereal to have been widely dispersed through this area as staple seems to have been foxtail millet (Setaria italica; Indon. jawawut), represented in Proto-Austronesian by the protoform *beCeng, having reflexes in Taiwan, through the Philippines, Central and East Indonesian up to Cendrawasih Bay in the North of Irian Jaya. The word was also borrowed into Non-Austronesian languages of North Halmahera.
The grain is first attested in Timor in a layer slightly younger than 1000 BC, and is the earliest cereal here (aside from Job's tears, Coix lacryma-jobii; Indon. jelai = enjelai, which occurs wild here and over most of Indonesia). that is about one millennium later than earliest finds of the domesticated pig on the island. Apparently, Austronesians proceeding further south from Luzon for the first thousand years did not have a cereal staple but tubers instead, and they are the first Austronesian setlers in Maluku, from among whom were the ancestors of the Oceanians. Only later, after having sufficiently learned to cultivate the cereals in tropical conditions, did Austronesians moving south from Luzon cultivate a cereal as staple, and this was foxtail millet. Rice in sufficient quantity was apparently only cultivated even later, around 500 BC in Sulawesi (Peter Bellwood) and the Malayan Peninsula (Janice Stargardt). Reflexes of *pajei are also represented till as far east as Cendrawasih Bay.
This leads to TWO CONCLUSIONS having relevance to our problem of the relationship between Maluku and Melanesia. Although rice was indeed cultivated in the Austronesian homeland in Taiwan perhaps as early as time of Proto-Austronesia, whereas the tubers were probably not, the first Austronesians to head for Oceania from Maluku were probably indeed growers of the tubers, and not of rice or another cereal. While a distinct local culture tradition, the Lapita culture, was developing and spread out in Melanesia, later newcomers arrived on the scene in Maluku, mingling with the first settled Austronesians (which were closestly relate to the Oceanians), apparently abandoning their language for that of the earlier settlers (for which reason no non-CEMP Austronesian languages are found in the area). These newcomers, however, brought the technology of cereal cultivation here, a culture feature not shared by the Melanesians.
It is possible, that an analogical process of introduction of a new staple crop also took place in Melanesia, though at a much later period. Apparently, contacts between Polynesians and Amerindians led to the introduction of the sweet potato (Ipomea spp., particularly Ipomea batatas; Indon. ketela, particularly ketela rambut) into Oceania. The crop seems to have been gradually propagated westwards through trade contacts, though not further than New Guinea by contact time. Propagation further west into Maluku and the Philippines seems to have been the result of activities of the Spaniards. Incidentally, the word "ketela" apparently entered Malay through Maluku and reflects the word "Castilian" (through something like ubi ketela "Castilian yams").
It is of course not surprising, that the peoples of the two regions experienced divergent developments and influences after separation. Nevertheless, it is in my opinion possible that even after they had separated, there were contacts at several occasions. However, this touches a controversial point in the reconstruction of Oceanic linguistic and culture (pre)history. From a linguistic point of view, the Oceanic language group, encompassing all Austronesian languages is a compact language group. That is to say, it has a distinct single common precursor, Proto-Oceanic, which can be identified with the language of the ancestral peoples which originally found their way from the South Halmahera / North Irian Jaya area to the Bismarcks region. This is generally understood to mean that there was only one Austronesian movement into Oceania, and the archaeologists too seem to be quite satisfied with this picture.
However, the compactness of the language group does not, in my opinion, exclude later incursions from East Indonesia. Thus, the compactness of the Romance language group with vulgar Latin as its unique common precursor is not impaired by the well-known fact, that Germanic tribes settled in Gaul and Iberia, taking over the Romance languages of the local, originally Celtic population, or that Slavic peoples settled in Roumania, taking over the Romance language of local Dacians. In the Pacific too, in my opinion, there were several subsequent movements from East Indonesia which advanced varyingly deep into Oceania.
The first of these movements, in my opinion, introduced regular long-distance navigation and construction of megalithist monuments such as hill-top and ridge fortifications and stone temples. They seem to have been a dynamic and expansive peoples with a relatively advanced social stratification, perhaps even at the threshold to state formation, and seem to have sweeped through Indonesia from the center in two directions: eastwards all the way to Polynesia, intermingling with Oceanians and taking up their language as they passed through the Solomons and the South Pacific, and westwards all the way to South India, partially also to Northeast India up the Ganges/Brahmaputra systems. Their word for "sailing boat for long-distance navigation", *peDaHu / *paDaHu, is well represented along the route route from the Bismarcks till Polynesia, as well as in Dravidian languages of South India. This suggests that they were the ones who introduced long-distance navigation to both areas. Stone constructions reminiscent of those of Polynesia in several respects can be found in East Indonesia (particularly in Nusatenggara), Java, Sumatra, and Northeast India (I reported on this in a paper in 1994, scheduled to appear in print this year, and in part in a 1994 publication).
Crucial is here perhaps, that megalithism has not so far been indicated as a particular feature of Lapita. If there had not been a second wave, then the quite sophisticated megalithist monuments of Polynesia would have to be assumed to be an Oceanic innovation. Resemblance to megalithist constructions in Nusatenggara, Java, Sumatra and Northeast India would have to be considered as coincidents. This is not altogether unlikely, because no exact comparative archaeological studies have been made so far. But the distribution of reflexes of *peDaHu / *paDaHu would then be rather surprising. Assuming, on the contrary, that there was this second wave, then Maluku and Melanesia experienced some common influences or developments quite some time after completion of the first split. But this involves features not unique to East Indonesia in the West, but also shared by Central and East Indonesia, and even Northeastern and Southern India.
The establishment of regular maritime communication between India and West Indonesia seems to have facilitated the introduction of first staple cereal into the later region, sorghum (Sorghum vulgare = = Andropogon sorghum; Indon. jawaras, Toba Batak jabaure) just before foxtail millet was introduced into Indonesia. This seems to be the reason why reflexes of *beCeng "foxtail millet" are not represented in West Indonesia. Instead, the name for sorghum, *zawaH (from Pali java, reflecting Sanskrit yava "barley"; reflexes in some languages in India have the meaning "sorghum") reflected e.g. in Malay as jawa (as in jawa-wut, jawa-ras), became generalized to mean "grain", including foxtail millet. Thus we can place the date of movement of the *peDaHu-sailors through Maluku before earliest cereals dated right after 1000 BC in Timor. This also means that they did not cultivate cereals either, but probably had tubers as staple.
Another possible movement into Oceania seems to have taken place at a later time, probably not before 700 BC, but certainly before the dispersal of Dongson-type bronze and the "ship-of-the-dead" cult throughout Indonesia between the 5th century BC and the 1st century AD), because this culture feature did not advance eastwards further than Irian Jaya, and thus serves to differentiate Maluku from Melanesia. The preceding movement seems to have introduced the lashed-lug plank-hulled boat to the Northern Solomon Islands (which seems to have been the furthest eastern extent of the movement). The construction principle of the boat is identical with that of the orembai in Maluku, and otherwise only still found with the Yamis of Botel Tobago Island (Southeast of Taiwan). It originated perhaps as a result of copying of the so-called "dragon boats" of mainland Daic peoples of Southeast Asia and Southeast China by Austronesians accustomed to the lashed-lug technique of the traditional Austronesian dugout-based so-called "five elements" boat. This is perhaps the same movement which carried a Mon-Khmer form for betel peper (Piper betle; Indon. sirih), i.e. *bVlu[] (cf. Old Mon se-blu') through Sulawesi (cf. Mandar buulu), Nusatenggara and Southeast Maluku (cf. Bimanese bulu, Wetar huru) to the St. Vitiaz languages Southwest of the Bismarcks (Siassi ful, Gedaged fu).
This was apparently the last movement into Oceania, contributing to common culture features between Maluku and Melanesia, happening before two major developments which brought in the metal age into East Indonesia, and thus constituted the FIRST ESENTIAL CULTURALLY ALIENATING DEVELOPMENT in the relationship between Maluku and Melanesia.