From the time that A. R. Radcliffe-Brown provided the first detailed description in 1926, the Rainbow Serpent culture complex has been regarded, either explicitly or implicitly, as the exclusive domain of Aboriginal Australia. An extensive survey of the ethnology1 of the rainbow fails to support this view, since elements of the same belief system appear in animistic societies around the world. These typically represent the rainbow as an enormous spirit snake that guards springs and other water sources when not in the sky and is particularly offended by the approach of menstruating women, while in other cultures rainbow and serpent have been decoupled, but both retain clear signals that they descend from a coherent and presumably ancient Rainbow Serpent culture complex. In those trparts of the world where agriculture first took hold, leading to urbanization, literacy, and the abandonment of an animistic worldview, the serpent survived as an independent entity that nomi life.Thi , geth wih h i x lu i y Tod deiie we know as the dragon.

Abstract This chapter offers a view into the world of the speakers of Proto-Malayo-Polynesian, their culture (houses, village structure), subsistence (agriculture, maritime technology) and social organisation (marriage, kinship relations). It addresses linguistic contributions to the culture history of the Austronesian-speaking peoples of Southeast Asia and Madagascar. Linguistic evidence for human prehistory is based on the reconstruction of vocabulary, which in turn derives from application of the comparative method, a set of procedures for inferring aspects of prehistoric languages from the traces left in their attested descendants. While these procedures are mainly concerned with how to infer the shapes of words, the chapter also refers to Benveniste (1969) and previous work by Blust for a discussion and illustration of the principles of semantic reconstruction, noting inferences about culture history that depend equally on the meaning of reconstructed words.

Modernization and conversion to world religions are threatening the survival of traditional belief systems, leaving behind only mysterious traces of their existence. This book, based upon extensive research conducted over a period of nearly four decades, brings scientific rigor to one of the questions that have always attracted human curiosity: that of the origin of the “dragon”. The author demonstrates that both dragons and rainbows are cultural universals, that many of the traits that are attributed to dragons in widely separated part of the planet are also attributed to rainbows, and that the number and antiquity of such shared traits cannot be attributed to chance or common inheritance, but rather to common cognitive pathways by which human psychology has responded to the natural environment in a wide array of cultures around the world.

From the time of their earliest reconstructable ancestor, Austronesian lan-guages have avoided morphemes that allow dissimilar labials as the onsets of successive syllables. What is of interest to phonologists is that this inherited morpheme structure constraint continues to hold on the word level in many (but not all) daughter languages. As a result of the extension of phonotactic restrictions from the level of the morpheme to the level of the word, lexical bases that are affixed in ways that violate this constraint show various repairs aimed at removing a marked sequence, most commonly one in which a labial -initial base drops the initial syllable when infixed with-um-. The range of attested repair options is surveyed in relation to claims made in earlier analy-ses, additional support is given for previously-recognized repairs, and two new functionally-equivalent avoidance strategies are described. Finally, although the most common repair option (pseudo nasal substitution) may be inherited from Proto-Austronesian, other structurally-distinct but function-ally-equivalent repairs apparently constitute a drift, namely a product of the continued operation of inherited structural pressures after language separation, a type of change that can be viewed as the diachronic counterpart of a syn-chronic conspiracy.

Although Formosan languages provide little support for Proto-Austronesian lexical bases of the shape CVNCVC, reflexes of such forms are common outside Taiwan, except in the northern Philippines, where they are relatively rare. This observation has produced two contrasting positions about the history of medial prenasalization in Austronesian languages. Although his higher-order subgrouping ideas were unclear, Dempwolff allowed medial -NC- in his "Uraustronesisch" (equivalent to modern Proto-Malayo-Polynesian), and a similar practice appears in later work by Blust, where Proto-Malayo-Polynesian was explicitly recognized. In opposing this view, Reid, in works over the past forty years, has claimed that -NC- developed after Northern Philippine languages separated from all other Austronesian languages outside Taiwan, implying that there is no Philippine subgroup. Evidence is presented that many Northern Philippine languages contain native lexical items with -NC-, a conclusion that is consistent with the reduced incidence of such forms in non-Northern Philippine languages such as Tagalog, and with clear evidence of a Philippine branch of the Austronesian family.

While a fair amount has been written on semantic change, semantic reconstruction is in some ways the last frontier in historical linguistics. Following principles laid down in past publications starting in 1987, I argue here that Dempwolff's classic reconstruction of “Original Austronesian” *guntiŋ ‘scissors’ is questionable, based both on records of the history of technology, and on evidence that *guntiŋ had another, more plausible meaning by at least PMP times (correlated with the Philippine Neolithic starting circa 2,200 BC). In particular, reflexes of this form (which appears as *sala-guntiŋ in some Philippine languages) refer to an X-shaped architectural structure used to support roof beams in traditional house construction. In addition, in languagesranging from the northeast Philippines to the Malay peninsula, it evidently designated a similar structure used to hold a spear in horizontal position prior to being triggered by an animal taking the bait in a *balatik spring-set spear trap used to take wild pigs. Both of these are features of Neolithic technology that was widely-shared by PMP times, and there is little need to assume that they were borrowed. Scissors, on the other hand, which share the same X-shape when opened, were extremely useful in hair-cutting when once obtained, and quickly acquired the name guntiŋ because of their shape. If they arrived from the Middle East, as seems likely, they would probably have first been acquired by Malays, who then passed them on (along with many other loanwords) to languages all over the Indo-Malaysian archipelago and the Philippines to their north.

This book documents an understudied phenomenon in Austronesian languages, namely the existence of recurrent submorphemic sound-meaning associations of the general form -CVC. It fills a critical gap in scholarship on these languages by bringing together a large body of data in one place, and by discussing some of the theoretical issues that arise in analyzing this data. Following an introduction which presents the topic, it includes a critical review of the relevant literature over the past century, and discussions of the following: 1. problems in finding the root (the "needle in the haystack" problem), 2. root ambiguity, 3. controls on chance as an interfering factor, 4. unrecognized morphology as a possible factor in duplicating evidence, 5. the shape/structure of the root, 6. referents of roots, 7. the origin of roots, 8. the problem of distinguishing false cognates produced by convergence in root-bearing morphemes from legitimate comparisons resulting from divergent descent, and 9. the problem of explaining how submorphemes are transmitted across generations of speakers independently of the morphemes that host them. The remainder of the book consists of a list of sources for the 197 languages from which data is drawn, followed by the roots with supporting evidence, a short appendix, and references.

At least three types of sound change (prenasalization of obstruents, lenition of obstruents, conversion of labial consonants to the corresponding labiovelars) are widespread in the Austronesian language family as sporadic innovations. What marks these off as different from more familiar types of irregularity is their repeated occurrence across hundreds of related languages, a phenomenon that can conveniently be called “pandemic irregularity.” All attempts to find an explanation for why pandemic irregularities occur in terms of possibly unrecognized affixation, conditioning, borrowing, or unfinished processes, have proven futile. In particular, it is stressed that pandemic irregularity in sound change is fundamentally different from “lexical diffusion”, and deserves to be recognized in its own right as a process that works against the general application of the regularity hypothesis.

The argument for a Philippine subgroup of Austronesian languages that has been presented by Blust is further strengthened through the addition of 320 new etymologies, amounting to an increase of about 25% over the earlier dataset. While the earlier publication aimed at comprehensiveness, this one adopts a more restrictive approach that virtually eliminates the likelihood of cognate distributions being the result of borrowing. The emphasis is thus on the quality of individual comparisons, rather than on the impact of an argument based, at least in part, on quantity. In addition, it clearly describes the foundations of the theory being defended, and provides an explicit discussion of method that lays bare certain misconceptions about the nature of historical linguistics held by some critics of the earlier proposal.

This squib describes three theoretically challenging phonological processes found in the Manide, Inagta Alabat, and Umiray Dumaget languages spoken on the large northern Philippine island of Luzon. These three well-documented processes do not conform to current theoretical expectations about what is a likely or even a possible diachronic process, although each is part of a larger, complex context of sound change which does conform to theoretical expectation. A brief background survey of vocalic changes triggered by voiced stops is given, followed by the puzzling changes that depart from this more general pattern.

For over a century anthropologists and folklorists have sporadically recorded a belief that one should not point at a rainbow, lest the offending finger become permanently bent, rot, be supernaturally severed, fall off, etc. In each case the belief was reported for a particular geographical region without apparent awareness of its presence elsewhere, and in no case was an explanation for this curious idea proposed. This paper documents what is called the “Rainbow Taboo” as a global phenomenon, found among peoples of quite varied cultural backgrounds, and it argues that the universality of the belief is a product of the interaction of two independent cognitive elements: an apparently innate sense that the rainbow is associated with the “other world,” and, secondly, a similar sense that pointing with the index finger is aggressive, and should not be used either in normal human interactions or more particularly against the supernatural.

Abstract:I am grateful to the four colleagues who have expressed their views on my argument for the reality of PPH, and I wish we could have had a larger number of contributors, with somewhat shorter and more relevant commentary in some cases to provide space for others.

Miri, an Austronesian language spoken in northern Sarawak, Malaysia, has two sets of vowel changes that are conditioned by voiced obstruents. In the first set, a last-syllable low vowel is fronted and raised to [e], or less commonly [i], if a voiced obstruent appears earlier in the word, while a penultimate low vowel immediately following the trigger is skipped. In the second, a high vowel in the final syllable undergoes breaking (diphthongisation) or lowering, depending upon specific conditions, unless there is a voiced obstruent anywhere earlier in the word. For both triggers and suppressors, this effect is cancelled by an intervening blocking consonant, which includes any nasal or voiceless obstruent except glottal stop. The challenge is to understand why voiced obstruents have this double function, acting as a trigger with low vowels and a suppressor with high vowels, given the lack of an a priori transparent relationship between low vowel fronting and high vowel breaking/lowering.

Dragons are portrayed in widely separated cultures as simultaneously male and female. While this trait may appear arbitrary, its global distribution implies that it is motivated by observations available to all humans. Many seemingly arbitrary traits of dragons are shared by conceptions of the rainbow, which is widely portrayed as having both male (primary) and female (secondary) arcs. This correlation of features in mythical and natural counterparts shows that the nineteenth-century doctrine of naturalism, which saw various elements of myth and folklore as reflecting features of the natural world, was valid despite excesses that eventually caused its abandonment.

Although a Philippine language group was tacitly assumed by most scholars for much of the twentieth century, more recent evidence-based attempts to determine the position of the languages of the Philippines within the Austronesian language family have generated an ongoing debate. In 1982, Lawrence A. Reid proposed an Austronesian family tree with five primary branches: Atayalic, Tsouic, Other Formosan, Bilic, and Amis-ExtraFormosan. Not only did this proposal make the Bilic languages of southern Mindanao extremely remote relatives of other languages in the Philippines, but it also asserted that the languages of northern Luzon (called "Outer Philippines") form a primary branch of a proposed Extra-Formosan group as against the residue ("Malayo-Polynesian"), leaving the idea of a Philippine subgroup in total disarray. Four years later Reid's position was challenged by David Zorc, who argued on the basis of 98 proposed lexical innovations that, apart from Sama-Bajaw, all languages of the Philippines form a genetic unit with the Sangiric, Minahasan, and Gorontalic languages of northern Sulawesi. More recently Malcolm Ross has also questioned ProtoPhilippines, holding that the languages of the Batanes islands between Taiwan and Luzon are an in situ continuation of the initial Austronesian settlement of the Philippines. These claims conflict with masses of counterevidence, which supports the reality of a Philippine group and implies that sometime after the initial phase of Austronesian settlement Proto-Philippines expanded at the expense of other related languages in the archipelago.