The cross correlation between the Hlai word for ‘to die’ and the Austronesian and Tai-Kadai words for ‘to scratch’


The word for ‘to die’ has been widely attested in the historical and comparative linguistics studies of Austronesian and Tai-Kadai for a long time. The proto-forms have been reconstructed as Proto-Austronesian *m-aCay, Proto-Tai *p.ta:jA and Proto-Kam-Sui *pjai1 (Austronesian Basic Vocabulary Database by Greenhill, S.J., Blust. R, and Gray, R.D. 2008). Except, Kra and Hlai have used the different words, instead with unknown source and without clear explanation provided so far. In the connection to this puzzle, it is the main purpose of the present paper in order to trying to answer the question but only the Hlai with very much appreciated to Peter Norquest on his study of A Phonological Reconstruction of Proto-Hlai 2007.

The present study has been conducted by determining the subsurface etyma of Indonesian disyllabic correspondence set, in this case called ‘monosyllabic root’ and then followed by the cross correlation between Austronesian and Tai-Kadai which may be different from the mainstream works. The term ‘subsurface etyma’ might sound strange in the linguistics field because whenever the term ‘etymon’ is applied it should mean that there is no any further unknown root remained. However, in Austronesian languages, there still be many core correspondences that have not fully been investigated, even though, in the past 15-30 years, many scholars, such as Robert A. Blust, R. David Zorc, John U. Wolff, E.M. Kempler Cohen, etc. have identified the vast area of monosyllabic roots. In Tai-Kadai, the sound-meaning correlated words have been recognized by some scholars since 1963 and lately by Navavan Bandhumedha in 2012, however, the monosyllabic root has never been identified before because the difficulty is that how can we isolate the monosyllabic root from the monosyllabic and sesquisyllabic languages.

Nevertheless, the problem could be eliminated by the strong support from the Austronesian monosyllabic roots and their subsurface etyma. As recently identified, Tai-Kadai have shared at least 32 monosyllabic roots with Austronesian of which 19 roots are tied with the previous identifications. They are called ‘cross-correlated monosyllabic roots’ that strongly support of the coexistence of Austro-Tai and considered as the parent of many basic disyllabic words sets in Proto-Austronesian as well as Proto-Tai-Kadai, such as numerals, body part terms, nature terms, living society terms, basic animals and basic verbs. The monosyllabic roots have reflected one basic rule of ancient languages that is a pair characteristic of male and female, dominated by *l-sound and *r-sound, respectively (see The monosyllabic roots of Austronesian and Tai-Kadai: a proof for Austro-Tai, Supat Charoensappuech 2017).    

The Hlai word for ‘to die’ has two distinctions but the interested one under the present paper is ɬa:w2 and ɗa:w2 which have been used by Bouhin as ɗa:w2 and by Ha Em, Lauhut, Baoting and Baisha as ɬa:w2. Peter Norquest has reconstructed Proto-Hlai *hla:wɦ and Pre-Hlai *la:wɦ, comparing to Weera Ostapirat’s Proto-Hlai *ala:uB. There are 6 correspondence points that can be correlated as follows:

The first correspondence is that Baoting has used ɬa:w2 in 3 words; ‘to die’, ‘spade’ and ‘take off’.

The second correspondence is that Peter Norquest’s Proto-Hlai *hla:wɦ and Pre-Hlai *la:wɦ are very similar to proto-form of word for ‘squirt’ as Proto-Hlai *hla:w and Pre-Hlai *la:w.

The third correspondence is that word for ‘spade’ and ‘take off’ have the same proto-form as Proto-Hlai *lja:wɦ and Pre-Hlai *Cila:wɦ of which are not far from the proto-form of words for ‘to die’ and ‘squirt’.

The forth correspondence, the most important one, is that words for ‘squirt’, ‘spade’ and ‘take off’ are considered as the subset of *raw, another cross-correlated monosyllabic root of Austronesian and Tai-Kadai. It has the main fundamental sense of ‘to scratch, carve and outline’ and it is paired with *law, one of the cross-correlated monosyllabic root who acts as male with the fundamental sense of ‘all gathered, mixed up and oldest source’.

In Tai-Kadai, this root is clearly identified in the area of Tai and Hlai and could be separated into two classes based on the step of working as follows:

1. ‘to scratch and carve’:

เกลา – klawA1              ‘to polish, to clear, to smooth, to scrape’ (it is a universal word in Siamese that has been using for such actions) 

เกา kawA1                 ‘to scratch’ (Tai (Pittayaporn Pittayawat 2009) - Siamese, Lungchow, Shangsi, Yay and Saek kawA1 and Proto-Tai *kaw A and it is supposed to be collapsed from *klawA or *krawA; Hlai (Peter Norquest 2007) - Bouhin (ka:w1), Ha Em and Lauhut ga:w1, Tongzha and Changjiang ga:w4 and Moyfaw ɣa:w1 and Proto-Hlai *ra:w

เสลา slawA1             ‘to carve’ (Siamese word; automatically, it means to the sleek and carve processes of any material until finish and is always used with a word for ‘sculpture’ สลัก slakD1, a subset of cross-correlated monosyllabic root *lak who has the fundamental sense of ‘male trait’) 

เปลาplawA1              ‘sleek wood, long tube’ (Siamese word; it is the old word and have the similar sense with เกลา – klawA1and เสลา slawA1)

เสา sawA1                ‘pillar’ (เสา sawA1 ‘pillar’ is the final product of to scratch, to carve, to scrape, to smooth, to sleek and to make shape, therefore, this monosyllabic word is considered as a native Tai-Kadai that should be collapsed from sysquisyllable, perhaps from *slawA or *srawA and has widely been used in Tai-Kadai, except Hlai, such as Tai - Siamese and Sapa sawA1, Bao Yen thɤwA1, Cao Bang ɬɤwA1, Lungchow and Shangsi ɬawA1, Yay θawA1 and Proto-Tai *sawA; Kam-Sui (Ilya Peiros 1998) - Then za:u2, Lingam Sui la:u1 and Mak za:u1; Kra (Weera Ostapirat 2000) - Gelao saA1, Lachi ʨiA1, Laha couB2 –t, Paha dʑhuuA1, Buyang θuuA1, Pubiao ʨauA1 and Proto-Kra *m-tȿuA)

เพลา – phlawA2           ‘shaft’ (Siamese word; it is represent the long tube-shaped that has been done through the same process as others)

เหลา – lawA1               ‘to sharpen, to chamfer’ (it is another universal word in Siamese that applies to sleeking work on a stalk, such as bamboo and rattan and followed by making a shape)

หลาว – la:wA1            ‘spear and to spear’ (it is used by Siamese in two kinds as the long arms and the action by preceding to the front; in Hlai, it could be compared to word for ‘squirt’, such as Ha Em, Lauhut, Tongzha , Zandui, Baoting, Moyfaw, Baisha and Yuanmen ɬa:w1 and Proto-Hlai *hla:w and Pre-Hlai *la:w)

za:w2                           ‘spade’ (Hlai - Bouhin, Ha Em, Lauhut, Moyfaw and Baisha za:w2, Tongzha, Zandui and Baoting ɬa:w2, Yuanmen tsa:w2 and Proto-Hlai *lja:wɦ and Pre-Hlai *Cila:wɦ; the sense that related with ‘to scratch and carve’ is not only by the application of material, it is also related by its wooden handle which was sleeked from the long tube or slender stalk)

ราว ra:wA2               ‘about, bar’ (Siamese; it has two duties as the rough estimation and as the string or row of thing and much associated with words for ‘agile’ เปรียว priawA1, ‘slim’ เพรียว <i

sup>hriawA2, ‘slender’ เรียว riawA2 and ‘strip’ ริ้ว riwC2) </i

สาว – sa:wA1              ‘to pull and to strip’ (in Siamese, even though it uses the same pronunciation with word for ‘bride’, it has a different sense and comes from a distinct root; the Hlai word for ‘to strip’ is divided into two classes; the first one, such as Bouhin, Ha Em, Lauhut, Moyfaw and Baisha row1, Tongzha row4 , Zandui low4, Baoting (law5) and Proto-Hlai *ɾu: and Pre-Hlai *C-ɾu: and the second one, such as Ha Em and Lauhut law3, Baisha1, Yuanmen 4 and Proto-Hlai *C-lu:ʔ; ‘to pull and to strip’ is the meaning that falls under the same process with ‘to scratch and to carve’)

za:w2                           ‘to take off’ (Hlai - Bouhin, Ha Em, Lauhut, Changjiang and Moyfaw za:w2, Tongzha, Zandui and Baoting ɬa:w2, Cunhua law5 and Proto-Hlai *lja:wɦ and Pre-Hlai *Cila:wɦ; the observations can be made that Bouhin, Ha Em, Lauhut, Moyfaw have used za:w2 in words for ‘spade’ and ‘take off’ whereas Tongzha, Zandui, Baoting have used ɬa:w2 in words for ‘spade’ and ‘take off’ and   the Proto-Hlai and Pre-Hlai words for ‘spade’ and ‘take off’ are even the same; ‘take off’ has the same meaning with ‘to pull and to strip’ off something)

ซาว sa:wA2              ‘to wash’ (Tai – Siamese sa:wA2, Cao Bang ra:wA2, Lungchow and Shangsi ɬa:wA2 and Proto-Tai *za:wA, it is a word that normally applies to washing the milled rice before cooking; Hlai – Cunhua saw3 and Moyfaw sa:w3 and Proto-North Central Hlai *ɕa:wʔ)

2.‘to make pattern, shape and structure’:

คร่าว khra:wB2         ‘outline and frame’ (in Siamese, it has two applications, one is used as a kind of cross bar structure, in particular, for the house wall and another is used as a rough boundary or rough structure or rough estimation which is similar to ราว ra:wA2 ‘about and bar’)

raw1                        ‘to sift’ (Hlai – Bouhin and Ha Em zaw1, Lauhut raw1, Tongzha and Baoting taw4, Zandui thaw4 and Proto-Hlai *ɾjəw and Pre-Hlai *Ciɾəw; the cross bar structure of คร่าว khra:wB2 ‘outline and frame’ and raw1 ‘to sift’ are in fact the part of each other)

za:w3                                         ‘barn’ (Hlai - Ha Em, Lauhut, Changjiang, Moyfaw and Baisha za:w3, Tongzha, Zandui and Yuanmen za:w6, Nadouhua zaw3 and Proto-Hlai *hja:wʔ and Pre-Hlai *ja:wʔ; a ‘barn’ that used for storage the harvested rice in the past is anticipated to be made from the bamboo structure of which, in general, is a kind of คร่าว khra:wB2 ‘outline and frame’ and with interested it is quite similar to Kra word for ‘den/nest’, such as Paha ðaauC1 and Pubiao θooC1 and Proto-Kra *trauC)         

In Austronesian, *raw can be identified in many disyllabic correspondence words in Indonesian. The examples are presented with its pair word from *law, if available, as follows:

derau                           ‘roar, loud, noise’ (such as the strong wind during heavy rain)

garau                           ‘hoarse’ (it is a sound that could be dominated by scratching; compared to galau ‘large place to gather, very dense without direction’)

kerau                           ‘bamboo basket’ (it is a strong evidence to show the process of scratching, sleeking and weaving of ancient material, such as bamboo as well as rattan) 

kirau                            ‘hard, rough and raw’ (compared to kilau ‘origin or source of sparkling and shining’)   

parau                           ‘hoarse and husky or to separate’ (hoarse and husky should be one of the ancient abrasive sound that related to scratching very well; to separate has a common ground with to take off or to strip and even to scratch; compared to palau ‘old mark or furrow’)    

raut                             ‘to scratch and make shape’ (it is a clear process of sleeking and weaving the product, such as basket, living material until the living structure; compared to laut ‘the mother sea’)

sarau                           ‘large basket’ (the meaning can well be extended to the structure like a barn or a house which is made from the bamboo or rattan)

serau                           ‘apart, aloof, wide pattern’ (it is the meaning that suits very well with to sift or to make a crossing structure)  

The examples given above have strongly expressed the fundamental sense of root *raw as ‘to scratch, to sleek and to make something’ of which points to the same root with Tai-Kadai. However, it could well be isolated from root *law, not like in Tai-Kadai who had been believed to shift the *r-sound to *l-sound in many words. The shifting between *r-sound and *l-sound in Tai-Kadai seems to be very obvious; the vast samples can be identified but do require help from the monosyllabic root of *l-initial and *r-initial who are a forebear pair. The distinction between *l-initial and *r-initial roots could, at least, be extracted from the correspondence set in Indonesian languages who have well preserved the disyllabic form as well as distinct sound and meaning. Then, the sub-grouping of *l-initial and *r-initial roots in Tai-Kadai should rather set the priority to its base meaning than l-sound or r-sound.                   

Furthermore, the root *raw does not only present in Indonesian, it has even embedded in the other Austronesian, in particular, word for ‘to scratch’. Many groups of Formosan have used word for ‘to scartch’, such as:

Basay L04 halaw

Kavalan kaɣau, qa:ɣáw, qaʀaw

Pazih F69 kaikaxáu

Saisiyat koma'ka'aw, kakaaw, kaLkaLaw

Saaroa kara:ro, karaaru.

Many groups of Malayo-Polynesian have also used this word in the similar sound and meaning as follow:

Grater Barito: Mapun and Bajo kakayaw, Dayak Ngaju manqqayau

Cham garao

Malayic:          Urak Lawoiʹ garu, Iban garuʔ, Kerinci gaewʔ, ŋaew, Banjarese Malay garu, Melayu menggaru

North Borneo: Belait mariow, Kelabit (Bario) ŋaro, kenyah (Long Anap) ñemayaw

Philippine:      Buhid kagáw, Bikol (Buhinon) kaˈɡaw, Hanunoo and Iraya kagaw.

Interestingly, the Proto-Malayo-Polynesian and Proto-Austronesian have been reconstructed as disyllable *kaʀaw. The form, even though the medial is , as well as the meaning of this proto-form have highly corresponded with *raw and it is also can be correlated with Tai-Kadai เกลา – klawA1 ‘to polish, to clear, to smooth, to scrape’ and เกาkawA1 ‘to scratch’. Therefore, it should not be overstatement that word for ‘to scratch’ *kaʀaw is a part of cross-correlated monosyllabic root *raw.

It has come to the conclusion between the line that *raw has long been existed in the Austronesian and Tai-Kadai as a cross-correlated monosyllabic root. It has the general sense of ‘making the product from the long tube-shaped material by the process of scratching, abrasive, scraping, craving, sleeking, sharpening, cleaning, washing, weaving and making a pattern’. Thus, needless to say that the Hlai words for ‘squirt’, ‘spade’ and ‘take off’ are a part of the root *raw. Apart, the present paper is one of the good samples for applying the term ‘subsurface etyma’ to the studies of these two phylums.

Moreover, it is one of the good indications which reflects the living style of people in the ancient time. Those were the forebear people who had the knowledge in producing the final product material from the raw material, such as bamboo and rattan through the process of scratching, carving and making structure. The final product could be ranged from the small cultural material until the large living structure. Upon the dispersed time, the root *raw with its knowledge was carefully carried on the back and shoulders of the immigrants to the new settlements and has been inherited from generation to generation until the present.

The fifth correspondence is that by nature, the dead body of any life could be identified by a simple practice, such as no breath taking in and out, motionless, stiff and lay down such like a dead wood. Hence, the dead body is very similar to a spade handle or a spear or a bamboo stalk of which is exactly pointed to the Hlai word for ‘to die’ ɬa:w2 and ɗa:w2 with Proto-Hlai *hla:wɦ and Pre-Hlai *la:wɦ.

The last correspondence is quite surprised because it is found that the Hlai word for ‘to die’ has well been preserved in Siamese and Lao in the words for ‘to stay still’ เซ่า sawB2, ‘inactive or stagnant’ เซา sawA2 and ‘to die’ ด่าวda:wB1, for example:

            เซ่าปาก – sawB2 pa:kD1          ‘please shut your mouth up’

            ซบเซา – sopD2 sawA2              ‘it is very quiet’

            ด่าวดิ้น – da:wB1 dinC1             ‘it is going to die’

Based on 6 correspondences as demonstrated in above, the conclusion is that the Hlai word for ‘to die’ could be correlated with much confident to the Austronesian and Tai-Kadai words for ‘to scratch’ through the cross-correlated monosyllabic root *raw under the fundamental sense of ‘to scratch, to carving and to make shape’. It is one of the ancient roots, supposed to exist since a time much earlier than the forming of Austronesian and Tai-Kadai; perhaps, earlier than the forming of the first generation disyllabic word. Therefore, the Hlai word for ‘to die’ is somehow nothing else, just a long naked and dried wood.

Supat Charoensappuech

Chanthaburi 22 August 2017

บันทึกนี้เขียนที่ GotoKnow โดย  ใน รอยไถของจักรวาล

ความเห็น (2)

IP: xxx.158.57.28

Thank you.

So 'to die' is 'to become stiff like a piece of dead wood'?

My ipa code reading is quite limited and I have problems working out the tones of words (in ipa code), I would appreciate a brief explanation on how to pronouce those words ;-)


Yes, 'to die' in Hlai languages is to express the sense of 'a long dead wood'. In the paper, there are two kinds of symbols; Sinitic sytem and Western system. A = low tone, B = mid tone, C = high tone and D = dead tone; 1 represents the voiceless (mid and high class) consonants and 2 represents the voiced (low class) consonant.

Sinitic Pattern 1 2 3 4 5 6 7 8 9 10
Western Pattern A1 A2 C1 C2 B1 B2 DS1 DS2 DL1 DL2