Spoken Language Iconicity: an Articulatory-Based Analysis of 66 Languages

Home , Cherokee language, Samu language, Sandawe language, Sierra Popoluca

國立交通大學

外國語文學系暨外國文學與語言學碩士班

碩士論文

口語象似性: 66 種語言構音為本的分析

Spoken Language Iconicity: An Articulatory-based Analysis of 66 Languages

研究生：朱易安

指導教授：賴郁雯博士呂佳蓉博士

中華民國一百零七年二月口語象似性: 66 種語言構音為本的分析 Spoken Language Iconicity: An Articulatory-based Analysis of 66 Languages

研究生：朱易安 Student: Ian Joo 指導教授：賴郁雯 Advisor: Yuwen Lai 呂佳蓉 Chiarung Lu

國立交通大學外國語文學系暨外國文學與語言學碩士班碩士論文

A Thesis Submitted to Graduate Program of Foreign Literatures and Linguistics Department of Foreign Languages and Literatures National Chiao Tung University in partial fulfilment of the requirements for the Degree of Master in

Foreign Literatures and Linguistics

February 2018

Hsinchu, Taiwan

中華民國一百零七年二月口語象似性: 66 種語言構音為本的分析指導教授：賴郁雯博士學生：朱易安呂佳蓉博士國立交通大學外國語文學系暨外國文學與語言學碩士班

詞彙的聲音跟語意是否具有相關性? 例如，圓唇母音會不會在代表圓形的事物的字詞中比平唇母音更常出現? 本論文調查 66 種沒有譜系關係的語言，選取 100 個基本詞彙分析，研究顯示 36% 的字詞與特定聲音有很強的正或負相關。這個結果與前人研究中的感知實驗和/或其類型研究吻合。例如：表達圓形的身體部位的詞素 (‘臍’，‘脖’， ‘乳’，‘膝’) 常有圓唇母音，本文推測肇因於圓唇和圓型之間的構音類似性。代表口腔運動的詞素跟類似的構音動作有相互關係 (例如 ‘吹’ = 唇字音，擦音，圓唇母音)。這些關係再次證明口語裡的語音與語意的匹配不是隨機的，而呈現某種程度的象似性。

關字: 象似性，類型學，聲音象，語音學，語意學

i Spoken Language Iconicity: An Articulatory-based Analysis of 66 Languages Advisor: Dr. Yuwen Lai Student: Ian Joo Dr. Chiarung Lu Graduate Program of Foreign Literatures and Linguistics National Chiao Tung University

Abstract

Is the phonetic form of a lexical item related to its semantic value in any way? For example, do rounded vowels occur more frequently in lexical items that refer to round objects than unrounded vowels do? Based on a wordlist of 100 basic lexical terms from 66 genealogically unrelated languages, I have investigated whether each term tends to be represented by morphemes containing or not containing certain sounds. The results show that 36 out of 100 meanings show preference or dispreference for certain sounds, many of these correlations sup- ported by previous experimental studies and/or cross-checked by other typological studies. For instance, morphemes that refer to round body parts (‘navel’, ‘neck’, ‘breast’, and ‘knee’) frequently bear rounded vowels, possibly due to the articulatory similarity between lip rounding and round shapes. Morphemes that signify oral actions correlate with the similar articulatory gestures (e. g. ‘to blow’ = labial consonants, fricatives, and rounded vowels). These correlations once again demonstrate that the mapping between sound and meaning in spoken languages may not be entirely arbitrary but can be, to some degree, iconic. Keywords: iconicity, typology, sound symbolism, phonetics, semantics

ii Acknowledgement

“That’s a lot to do for a Master’s thesis,” or so I’ve heard from several people when I told them about this thesis project. It was indeed an overwhelming task for an M. A. student to collect thousands of morphemes from dozens of languages to conduct an analysis that attempts to argue against one of the most broadly accepted principle of linguistics - the arbitrariness of signs. It was only with the help of people from all over the world - many of whom I have never met - that this challenging work was finally made into a thesis. I first thank my advisors, Yuwen Lai (National Chiao Tung University) and Chiarung Lu (National Taiwan University). Prof. Lai gave me detailed comments about the phonetic side of my thesis, while Prof. Lu provided me helpful insights into the field of cognitive linguistics. As this thesis is a phonetic and a semantic study, the two advisors with their different expertise were the best of both worlds for me. 我非常感謝兩位老師的指導。 Other people I must thank are those who provided me the data of many minority languages: Adam Boyd, John Sullivan, Claudine Chamoreau, Arok Wolvengrey, Lynda de Jong Boudreault, Richard Scaglion, Lise M. Dobrin, Bonny Sands, Minik Jeremiassen, Jill Riepe, Jesse Pryor, Karie Pryor, Cliff Goddard, Francesc Queixalós, Andy Minch, and Gatot (a native speaker of Galela). Syukuru dala-dala, Gatot. I would also like to thank my friends: Alan Nielsen, who encouraged me to use more rigid statistical methods, and Thomas Van Hoey, who thoroughly proofread my thesis. Thank you Alan, dank je wel Thomas. And lastly, how can I not thank my mom, my dad, and my brother who always trusted that what I was doing meant something. This thesis is based on their years of financial and emotional support and is thus dedicated to them. 엄마, 아빠, 상진아, 고마워.

iii Table of Contents

...... i Abstract ...... ii Acknowledgement ...... iii Table of Contents ...... iv List of Figures ...... vi List of Tables ...... vii 1 Introduction ...... 1 2 Literature Review ...... 3 2.1 Sign language iconicity ...... 3 2.2 Spoken Language Iconicity ...... 5 2.2.1 Ideophones and onomatopoeia ...... 6 2.2.2 Experimental Studies ...... 7 3 Research Question ...... 11 4 Previous Research ...... 13 4.1 Spatial Deixis ...... 13 4.2 Pronouns ...... 14 4.3 Swadesh List ...... 15 4.4 Diachronic iconicity ...... 19 5 Methodology ...... 23 5.1 Wordlist ...... 23 5.2 Sample Languages ...... 25 5.3 Database ...... 30 5.4 Transcription ...... 34 5.5 Classification ...... 34 5.6 Proportion Calculation ...... 39 5.7 Statistical Analysis of Correlations ...... 40

iv 6 Results and Discussion ...... 41 6.1 Positive and Negative Correlations ...... 41 6.2 Comparison with Blasi et al...... 44 6.3 Possible Motivations ...... 48 6.3.1 Mouth ...... 48 6.3.2 Shape ...... 48 6.3.3 Size ...... 48 6.3.4 Texture ...... 49 6.3.5 Deixis ...... 49 6.3.6 Verbal signs ...... 50 7 Conclusion ...... 51 References ...... 53 Appendix A The morphemes ...... 66 Appendix B Lexical sources and transcription method of each language ...... 158

v List of Figures

2.1 LIS sign for ‘napkin’...... 4

3.1 ASL sign for ‘sleep.’ Retrieved from http://www.handspeak.com on 2017/06/03. 12

5.1 Distribution map of sample languages. Longitude and latitude of each language were retrieved from Glottolog 3.0 [1]. Map created with Microsoft Excel 2016. 30

vi List of Tables

2.1 English and Vietnamese onomatopoeic expressions that resemble each other. . . 7

4.1 Concepts that show significant phonosemantic correlations, found by Blasi et al. 17 4.2 Concepts where Blasi et al. did not find any significant phonosemantic correlations 18

5.1 The LJ List...... 25 5.2 The 66 sample languages...... 27 5.3 Specifications for some of the LJ List terms...... 33 5.4 Phonemes that exist in the inventory of at least 70% of 2155 languages. . . . . 35 5.5 Phonemes and classes that exist in more than 70% of the languages...... 36 5.6 Common classes ...... 36 5.7 IPA characters of each class...... 38 5.8 The phonosemantic correlations represented by the p-values at different FDRs. 40

6.1 Positive and negative correlations...... 41 6.2 Comparison between the results of the two typological studies ...... 44

A.1 Spanish - Mandarin Chinese - Yoruba ...... 66 A.2 Egyptian Spoken Arabic - Javanese - Telugu ...... 70 A.3 Turkish - Japanese - Vietnamese ...... 74 A.4 Thai - Korean - Dholuo ...... 78 A.5 Hungarian - Chuanqiandian Cluster Miao - South Bolivian Quechua ...... 82 A.6 Peripheral Mongolian - Kabardian - K’iche’ ...... 86 A.7 Paraguayan Guarani - Georgian - Enga ...... 90 A.8 Eastern Huasteca Nahuatl - Central Aymara - Meqzuital Otomi ...... 95 A.9 Wayuu - Basque - Ngäbere ...... 99 A.10 Highland Totonac - Khoekhoe - Galela ...... 103 A.11 Mapudungun - Western Highland Purepecha - Woods Cree ...... 108 A.12 Navajo - Highland Popoluca - Ambulas ...... 113

vii A.13 Mískito - Shuar - Northern Emberá ...... 117 A.14 Bukiyip - Northwestern !Kung - Greenlandic ...... 121 A.15 Burushaski - Sentani - Terei ...... 126 A.16 Macushi - Ap Ma - Páez ...... 130 A.17 Wichí Lhamtés Vejoz - Sandawe - Xibe ...... 134 A.18 Toba - Ticuna - Kaingang ...... 138 A.19 Pitjantjatjara - Guahibo - Shipibo-Conibo ...... 142 A.20 Yanomamö - Tucano - Warao ...... 146 A.21 Awa-Cuaiquer - Mai Brat - Piaroa ...... 150 A.22 Amanab - Choctaw - Cherokee ...... 154

B.1 Lexical sources and transcription method of each language...... 158

viii Chapter 1 Introduction

I often have this strange and penetrating dream of an unknown woman, whom I love, and who loves me […] Is she brown-haired, blond, or redheaded? I do not know. Her name? I remember that it was soft and resonant Like those of the loved ones that Life has banished. –Paul Verlaine (1844-1896), “My Familiar Dream”

In the part of Verlaine’s poem cited above, the poet’s memory is –at first sight –logically impossible, but intuitively appealing. He only vaguely recalls the image of his dreamed lover and does not even remember her name. Yet he does maintain a vivid impression of what her name sounds like, which he describes as tragic and beloved. But what could be so “soft” about a name? Saussure [2] provided modern linguistics the principle of the arbitrariness of signs. His model of a linguistic sign was a binary connection between the signifier (the phonological concept) and the signified (the semantic concept). The association between the signifier and the signified is arbitrary and socially determined: a name is unrelated to the nature of the named. In Saussure’s view, the poet’s memory would make no sense. There could be nothing that makes a name “soft,” let alone “loved” or “banished.” Most of us, however, would agree that some names can sound soft and gentle, like Larry or Molly, while others may sound tough and ener- getic, like Kate or Kyle. If we look beyond the Saussurean framework, we see that the signifier-signified association is in fact not always fully arbitrary. Iconicity, the similarity between form and meaning, influ- ences human languages in diverse aspects and to various degrees. Many experimental studies, to be presented in the next section, confirm that certain linguistic sounds (like /l/ or /m/) can indeed be perceived as “soft” sounds and others (such as /t/ or /k/) as “hard” sounds. And such

1 correlations between linguistic form and meaning are well reflected in everyday words we use, as the present thesis will demonstrate. Section 2 will briefly guide the reader to the field of linguistic iconicity. Section 3 posits the research question of the present study and section 4 introduces previous research related to that question. Section 5 presents the methodology employed in my analysis, and Section 6 the results and the discussion. Section 7 is the final conclusion, and Section 8 the appendix.

2 Chapter 2 Literature Review

There have been constant challenges to the premise of the arbitrariness of signs. There is evidence that in our languages, sign or spoken, there exists some resemblance between the signifier and the signified. Not only is this resemblance evident in natural languages, it is also visible in experiments that demonstrate the perceptual bias towards the association between certain names and certain concepts.

2.1 Sign language iconicity

Sign languages are strikingly and undeniably iconic. Pietrandrea [3] observed that in the Italian Sign Language (lingua dei segni italiana, LIS), “[f]ifty percent of the occurrences of handshapes are motivated by an iconic association between their form and the portion of meaning they convey …Sixty-seven percent of the occurrences of body locations are motivated by an iconic association between the locations and the portion of the meaning they convey[.]” As a case of a sign whose handshape and location are both iconically motivated, consider the LIS sign for ‘napkin’ shown as figure 2.1, where “the location of the sign conveys the meaning mouth, and the movement conveys the meaning to rub.” This evident iconicity, arguably observable in any natural sign language, is strong counter- evidence against the idea of complete arbitrariness of linguistic signs. Sign language iconicity and spoken language iconicity are equally relevant to the broad question whether linguistic forms and meanings are interrelated (similar to how sign language phonology and spoken language phonology are equally relevant to the questions in phonology). A cross-linguistic observation of sign languages also hints to us that an iconic representation is, to some degree, conventional, because there can exist many different ways to iconically depict a given referent. Consider Klima and Bellugi’s [4] remark on how three sign languages– American Sign Language (ASL), Danish Sign Language, and Chinese Sign Language–express

3 Figure 2.1: LIS sign for ‘napkin’.

‘tree’ entirely differently, yet all iconically.

Consider the ASL sign for ‘tree.’ …[I]t is made with the forearm upright, the hand spread wide, and a twisting of the wrist and forearm. One could say that the upright forearm represents the trunk, the outstretched hand represents the branches, and the twisting motion represents the branches moving in the wind. In Danish Sign language the sign for ‘tree’ …differs in all details from the ASL sign, and yet it too is iconic: the two hands symmetrically outline the rounded shape of a tree’s top and then outline the shape of the trunk. The sign in Chinese Sign Language … is yet again different but still iconic: the two hands symmetrically encompass the shape of a tree’s trunk and move upward. Though the signs in these three languages are entirely distinct, both in the characteristics of tree they represent and the ways these are expressed in forming the signs, the signs are all iconic. (pp. 21-21)

The fact that linguistic signs for the same referent are different from each other does not rule out the fact that they may be iconic in different ways. Just as an artist has endless choices when illustrating a tree, a language has diverse ways to iconically express ‘tree’.

4 2.2 Spoken Language Iconicity

If sign languages are iconic, then is it really plausible that spoken languages are not? Ac- knowledging that sign languages are natural human languages with a modality different from that of spoken languages, it makes little sense to suppose that languages of one modality are iconic and those of another are not. The disregard for linguistic iconicity is simply a side effect from the perspective bias towards spoken languages, mainly because spoken languages are far more studied than sign languages are. One reason for this disregard is that spoken modality is generally more subtlly iconic than sign modality is. Taub [5] has pointed out that sign languages are more iconic than spoken languages because “most phenomena do not have a characteristic noise to be used in motivat- ing a linguistic form”, and “three dimensional spatial relationships, so crucial to language in many ways …cannot be represented iconically using the one-dimensional sequential medium of sound.” Although I agree with her the spoken languages are less iconic than sign languages, I do not think that sound is the only medium for spoken languages to iconically express a given concept. The sound is the medium of the speech perception. But for the speech production, we employ a broad portion of our body and go through a broad array of embodied experience. Like signers use their hands, arms, and face to articulate signs, speakers also use a whole array of speech articulators (including lips, tongue, teeth, jaw, oral cavity, nasal cavity, uvula, and so on) to express and deliver speech, and part of this embodied experience may bear an iconic resemblance to the concept to be expressed. For example, if a rounded vowel has an iconic correlation with a round shape, it may make more sense to think that the shape of the speaker’s rounded lips is iconically related to the image of a round shape than to believe there is something inherently “round” about the auditory characters of a rounded vowel. Iconicity emerges jointly from both the production and perception of human physical expression. As a non-linguistic example, consider the smile, which is a universal human expression of happiness. In the eyes of the perceiver, there is nothing inherently happy about the curvy shape of the lips. What makes a smile happy is that the smiler feels happier when smiling, i. e. con-

5 tracting the cheek’s zygomaticus major muscle, which elevates the feeling of amusement when contracted [6]. The perceiver, then, perceives the smile as a sign of happiness only because s/ he shares the embodied experience of the happy feeling of the contracted zygomaticus major. Likewise, it is possible that the perceiver of a rounded vowel perceives it to be related to a round shape because s/he shares the experience of rounding his/her lips. Sign and spoken modalities are both inherently iconic, albeit to different degrees. Again using the metaphor of an artist depicting a tree, the medium the artist employs limits how il- lustrative s/he can be. By using a colorful medium, such as oil paint, the artist can illustrate whether the leaves are dark or bright, whether the trunk is smooth or rough, and numerous other details. But if the artist is only given a pencil for his/her task, the details the s/he can convey are generally far more limited. Being generally more limited, however, does not imply that it is more limited in all aspects: just like oil paint can illustrate what a pencil cannot, so can a pencil more catch the details that paint oil fails to express, such as the fine wrinkles of the trunk. Like- wise, even though sign languages generally have more freedom to colorfully express meanings, spoken languages can often iconically express meanings in a way sign languages cannot (the most obvious case being the expression of aural concepts like onomatopoeia). Below, I will introduce previous studies that show how spoken languages are iconic.

2.2.1 Ideophones and onomatopoeia

An example of spoken language iconicity are ideophones –a lexical class that depict sen- sory images found in some languages (e. g. Japanese pika-pika ‘shiny’). Dingemanse et al. [7] showed that even without prior knowledge, naïve participants can guess the meanings of ideophones just listening to how they sound like. They gave Dutch speakers correct and incor- rect translations of ideophones from five languages and asked to choose the right one, and the participants made correct choices to a higher-than-average level. If ideophones fully arbitrarily represented their meanings, there would have been no way for participants to correctly guess their meanings. This shows that ideophones iconically represent their meanings to some degree. Another example of spoken language iconicity is onomatopoeia. Although English (bow wow), Korean (/mʌŋ mʌŋ/), and Japanese (wang wang) all express the sound of a dog barking

6 iconically, they do so differently. None of these is directly imitative of the sound a dog makes, but all bear sufficient resemblance to the referent to be easily recognized as iconic . Nguyen et al. [8] have illustrated how the onomatopoeic items expressing animal cries in two totally unrelated languages–English and Vietnamese–strongly resemble each other. Table 2.1 shows their examples, the left column being the English onomatopoeic words and the right column their Vietnamese counterparts.

Table 2.1: English and Vietnamese onomatopoeic expressions that resemble each other.

English Vietnamese (Tones not transcribed) coo gù /ɤʊu/ cheep chiếp /ciep/ hoo cú /kʊu/ cheep chíp /cip/ squawk quác /kwak/ moo bò /ɓɔ/ quack quạc /kwak/ caw quạ /kwa/ cuckoo cúc cu /kukp kʊu/ mew miu /miʊ/ meow mi-ao /miaʊ/

Considering the genealogical, areal, and historical unrelatedness between English and Viet- namese, it is unlikely that these highly similar words are of the same origin. Their similarity, therefore, must be due to their inherent similarity to the sounds they refer to. Even Saussure himself does not deny the sound-meaning resemblance of onomatopoeic words, even though he stresses that they are small in number, conventionalized throughout regu- lar sound change, and “jamais des éléments organiques d’un système linguistique (never organic elements of a linguistic system)” [2, p. 101] (he does not make clear what he means by “organic elements”). His principle of the arbitrariness of the sign is thus not an inviolable rule but only a general tendency.

2.2.2 Experimental Studies

A large portion of experimental phonosemantic research has investigated the sound-symbolic perception of pseudoword stimuli. Many experimental studies employed what we may call the

7 “sounds-like” methodology, where participants are asked to intuitively judge if a nonsense name “sounds like” a given concept.

Size

One of the most extensively studied phonosemantic phenomena is size-sound symbolism, also known as the magnitude symbolism. Sapir’s [9] experiment discovers that most participants, when given the two names mil and mal and informed that they both refer to tables, judge mil to be referring to a smaller table and mal to a larger one. If a evokes the image of larger size and i that of smaller size, which auditory or articulatory characteristics of these two vowels are relevant for the difference in perception of size? Shinohara and Kawahara [10] showed that native speakers of four different languages (En- glish, Chinese, Japanese, and Korean) generally associate high and front vowels with the smaller size. They propose that there are two possible explanations for this phenomenon: articulatory and auditory. The articulatory explanation would be that the enlarged size of oral cavity when articulating low and/or back vowels is analogically associated with larger size. The auditory explanation would be Ohala’s hypothesis of Frequency Code [11], which predicts sounds of lower frequency (fundamental or formant) denote larger size because the lower frequency is correlated with large animals with large resonators. But as Shinohara and Kawahara mentioned, if all high frequencies are associated with smaller size, it is unclear why low F1 is shown to be associated with a smaller size in their results. In Shinohara and Kawahara’s personal communication with Ohala, mentioned in their paper, Ohala pointed out that the articulatory hypothesis alone cannot explain why high F0 is associated with smallness since F0 has nothing to do with the size of the oral cavity. F0 was indeed shown to be correlated with size in LaPolla’s experimental study [12]: native speakers of Mandarin, when given nonsense syllables, assigned those with high tone (/55/) with meanings related to smallness and those with falling tones (/51/) with meanings related to largeness. On the other hand, even if F0 as tonal pitch plays a role in magnitude symbolism, it is evident that it plays no significant role as vocalic intrinsic pitch since high vowels have higher intrinsic F0 than low vowels [13] but high back rounded vowels evoke the perception of large size. The

8 best conclusion to arrive at this point seems to be that both articulatory and auditory mecha- nisms are at work. In other words, both oral cavity size and fundamental frequency influence the perception of size, and the former is prevalent in vowel-size correlation, whereas the latter prevails in tone-size correlation. Relatively little research has been done regarding which consonants sound larger. The abovementioned Shinohara and Kawahara’s study [10] found that voiced consonants sound larger to participants as opposed to voiceless ones, which they assume to be due to the lowered F0 of the vowels adjacent to voiced obstruents. LaPolla’s [12] experiment showed that Mandarin native speakers judged nonsense syllables with acute consonants to be referring to concepts related to smallness and those with grave consonants to concepts related to largeness. Acute consonants refer to what Jakobson [14] classifies as “dental” consonants (including alveolar consonants) and “palatal” consonants (including palato-alveolar consonants), and grave consonants to labial and velar consonants. In other words, acute consonants are roughly equivalent to coronal consonants plus palatal consonants, and grave ones to labial and dorsal consonants minus palatal consonants. Klink [15] found that participants from an American graduate school judged a motorcycle named Valp to be smaller (as well as faster and lighter) than the one named Galp. Since /v/ and /g/ differ from each other in both place and manner of articulation, it is difficult to tell which aspect caused the difference in size perception. We have two possible interpretations: 1. stops sound larger than fricatives; and/or 2. velars sound larger than labiodentals. Considering that both velar and labiodental consonants are grave sounds, the first interpretation seems more likely.

Shape

Another well-studied topic is the correlation between sound and shape, also known as the Takete-Maluma Phenomenon. Köhler [16] noted that when he showed people a picture of a curvy shape and a spiky shape and asked them which shape sounds like maluma and which shape like takete, most of them chose maluma for the curvy and takete for the spiky. Following Köhler’s remark, why and how this happens has been studied extensively. Fort

9 et al. [17] gave participants (French native speakers) a picture of a round shape and another one of a spiky shape with diverse pseudoword stimuli. participants “felt that” the pseudowords with the consonants /b/, /ʃ/, /d/, /l/, /m/, /n/, /s/, /p/, /ʒ/, and /g/ referred to the picture of a round shape, whereas those with /k/, /t/, /v/, /f/, and /z/ did not. We see that bilabial, nasal, and liquid consonants are linked to the round shape, whereas labiodentals are not. As for the vowels, the participants associated /i/ and /e/ with the spiky shape and /o/ and /u/ with the curvy one. The vowels only had marginally significant effect when they were paired with /t/, /s/, and /z/, and they had no significant effect when paired with /b/, /m/, or /l/. This could be because the association between curviness and /b/, /m/, and /l/ is so strong that it blurs the effect of the vowels. Fort et al. also remarked that the mean bias caused by front rounded vowels (/y, ø/) is in between that caused by /i, e/ and that caused by /o, u/. This suggests that both backness and lip rounding motivate preference for curviness. In sum, the preference for curviness is most strongly influenced by certain consonants, namely bilabial and liquid consonants, and also (to a lesser degree) by rounded vowels, especially back rounded vowels. This preference is apparently not culturally learned but innate: Ozturk et al. [18] found that even 4-month-olds react to sound-shape congruency and incongruency differently. When they were given “incongruent” combinations of sound and shape–listening to bubu when shown the spiky shape or listening to kiki when shown the round shape–they stared at the shapes longer than when the combinations were “congruent,” i. e. bubu with the round shape and kiki with the spiky shape. This may be because infants tend to stare at a strange sight longer than a normal one. Even for those of us older than infants, something that doesn’t seem normal tends to catch our eyes, as opposed to something that seems natural.

10 Chapter 3 Research Question

We have seen that there is extensive experimental evidence that certain sounds can evoke perceptions of certain size or shape. There follows the question whether these crossmodal correspondences are actually reflected in the lexicon of natural languages. In other words, if the sound of a rounded vowel evokes in our mind a round image, are rounded vowels actually more frequent than average in words that refer to round concepts? One may point out as counter-evidence that in natural languages, there are plenty of words without bilabials or rounded vowels that refer to round objects, or those without high front vowels that designate small things. The English word circle does not have any bilabial consonant or rounded vowel, and the word small has no high front vowel. Linguistic iconicity, however, however, is not a rule but a tendency. Even if some words meaning something round do not have any rounded vowels, it could still be the case that in general, words related to round shapes can have a higher-than-average ratio of rounded vowels. As an analogy, a single language with the verb-initial basic word order does not serve as counter-evidence to the statistical fact that most languages in the world (approximately 86%) have subject-initial basic word order, owing to the functional principle that discourse themes, usually a subject, tend to come first [19]. Only typological data that demonstrate subject-initial basic word order to be no more frequent than chance within a sizeable sample can work against the universal tendency for subject initiality. Likewise, the fact that a certain meaning (say, ‘round’) is not represented with certain sounds (bilabial consonants and rounded vowels) in a given language is not necessarily a counter-example to the tendency of that meaning to be represented with those sounds. One may argue the even though the spoken language onomatopoeia and ideophones are iconic, non-onomatopoeic and non-ideophonic words are not iconic at all. I strongly disagree with this “black-or-white hypothesis”. Although the onomatopoeia and ideophones may be the most prominent evidence for linguistic iconicity, they cannot be the only iconic constituent of

11 Figure 3.1: ASL sign for ‘sleep.’ Retrieved from http://www.handspeak.com on 2017/06/03. spoken languages. This is because we cannot neatly divide the natural language vocabulary into what is iconic and what is not. Such is evidently impossible when it comes to sign languages. Sign language lexicons are not divided into iconic and non-iconic signs: some signs are highly iconic, some are less so, and others are virtually not. See the ASL sign for ‘sleep,’ (fig. 3.1) which is only vaguely iconic. One could associate the closure of fingers to the closure of sleepy eyes, but no one without prior knowledge of ASL can guess what this sign means. As it is evident in sign languages, iconicity is gradient phenomena. Onomatopoeic expressions and ideophones are not 100% iconic (as they are also language-specific and conventional) and non-onomatopoeic and non-ideophonic words can also be iconic to various degrees, although they may often be less evidently so. The research question is thus the following: Are non-ideophonic and non-onomatopoeic lexicons of the spoken languages also iconic? That is, do spoken language lexical items that refer to concepts like ‘bird’ or ‘water’ also show iconic similarity to their meanings, similar to how ideophones and onomatopoeia do? For example, do lexical items that mean something round tend to have rounded vowels? The data I will present in this thesis says yes. But before I present my data, I will introduce some previous research on the iconicity of the general spoken language lexicons.

12 Chapter 4 Previous Research

4.1 Spatial Deixis

Several studies show that vowels are symbolically associated with spatial deictic terms. Proximal terms prefer vowels with higher F2, whereas distal terms prefer those with lower F2. Tanz [20] studied whether words for ‘here’ and ‘there’ in different languages tended to have certain sounds. She avoided the risk of biased sampling by choosing the following method:

With the aid of students in the Anthropology and Linguistics Departments at the University of Chicago and of residents at the International House, I gathered words for “here” and “there” in many languages […] They selected themselves at random for participation in the project by responding to an invitation posted on departmental bulletin boards to add to a list of words for “here” and “there” in another language. They did not know the purpose of collecting the list, and therefore could not have contributed words selectively to conform to the hypothesis. (p. 267)

Her results showed that in most of the cases, ‘here’ prefers /i/ and ‘there’ prefers /a/: there are cases where ‘here’ and ‘there’ contrast by vowel only (e. g. Kannada illi v. alli, Malay sini v. sana), where they differ in one syllable (e. g. Tulu /mulpa/ v. /alpa/, Turkish /burasí/ v. / buradá/), or where ‘there’ has an additional syllable (e. g. Arabic /huna/ v. /hunaka/, Woleloan /iga/ /igəla/). Woodworth [21], randomly sampling 26 languages, all from different language families, compared the vowels of deictic pronouns, place adverbs, and directional affixes. She found that in half of the cases, proximal terms have higher-F2 vowels than distal terms and only in one or two cases was it the opposite. In the rest of the cases, vowel comparison was either neutral or unclear.

13 Johansson and Zlatev [22] have investigated the iconicity of ‘this’ and ‘that.’ They selected 101 languages based on genealogical distribution, data availability, and areal distribution (Jo- hansson, p. c.). Their study shows that vowels in deictic terms are motivated mainly by their F2: higher F2 is related to shorter distance (‘this’) whereas lower F2 represents longer distance (‘that’). Other less prominent factors were vowel height (higher vowels tend to represent proximal terms, and lower vowels distal terms) and visibility (more visible articulations, like rounded vowels and labialized consonants, tend to represent distal terms, and less visible articulations, like velar consonants, tend to represent proximal terms).

4.2 Pronouns

1st and 2nd person pronouns tend to contain nasal sounds. Several cross-linguistic studies confirm this at least to some degree, although why exactly this is the case is not clear yet. Gordon [23] examined what sounds are the most commonly found in personal pronouns. From “genetically and areally diverse” (p. 118) 65 languages, he finds that 1st person pronouns, as compared to other pronouns, prefer /m/, especially in Eurasia. Gordon also finds that /t/ is common for 3rd person pronouns. Nichols and Peterson, [24] based on 173 areally and genealogically distributed languages, found that /n/ and /m/ are globally frequent in 1st and 2nd singular pronouns, although /m/ shows some areal heterogeneity.

In both first and second person pronouns, n is the most frequent nasal in the largest number of areas. This indicates that there is some cross-linguistic tendency to have n in pronominals and that this tendency makes itself felt in both persons, in various form categories, and in various parts of the world. On the other hand, the distribution of m, though also frequent and widespread, is less even. …m occurs in the majority or plurality in fewer continents than n does, and it is clearly associated with particular pronoun categories: first person in Eurasia and second person in the New World. (p. 351)

The study of Gordon and that of Nichols and Peterson both attest that /m/ is prevalent in 1st

14 person pronouns in Eurasia. Note, however, that the pronouns investigated in Gordon’s study are not the same as those studied in Nichols and Peterson’s. Gordon analyzed singular and plural pronouns, while Nichols and Peterson only analyzed singular pronouns. Furthermore, Gordon considered everything that “mark[s] the grammatical category of person” (p. 118) as a pronoun, thus including polite and vulgar forms, whereas Nichols and Peterson only considered pronouns of the neutral register. Despite these differences, it seems safe to say that Gordon’s results and Nichols and Peterson’s cross-check each other to some degree. Thus, nasals are preferred for 1st and 2nd person pronouns worldwide, but different continents prefer different nasals for each pronoun. One possible cause behind this combination of universal preference for nasal pronouns and continent-specific preference for specific nasals is what Nichols and Peterson name as “areal relatedness due to diffusion of phonosymbolic canons”. They further add that:

In personal pronoun systems, n and m can be said to mark different dimensions of a minimal deictic space. They do so as well in ’mama-papa’ systems (which are deictic but not shifters). Both the pronouns and the child-language kin terms use consonants phonosymbolically to structure deictic space; the phonosymbolic principles are macroareal (mama and papa, for instance, being distinctly western Eurasian forms); but the actual pronouns and kin terms themselves are not commonly borrowed. (p. 358)

Thus, there is the possibility that iconic patterns do not only independently emerge from individual languages but also areally diffused throughout neighboring languages.

4.3 Swadesh List

Blasi et al. [25] examined how the basic vocabulary of thousands of languages shows phonosemantic tendencies. The vocabulary they investigated were the Swadesh List, a list of 100 terms that refer to basic concepts such as human body parts, common animals, and natural substance. It was compiled by Swadesh [26] by revising a previously used longer list. The purpose of creating this list was to provide a tool for lexicostatistics, the comparison between two

15 languages or dialects that descended from the same ancestor (e. g. English and German) to see to what degree they share lexical cognates. For example, through the Swadesh List, we can see that English and German share more cognates (water–Wasser, fire–Feuer, man–Mann, etc.) than English and Russian do and that the former two are genealogically closer to each other than the latter two are. The Swadesh List was designed to be applicable to all human languages. Lexical terms that refer to what is not readily available in certain societies, such as ice, were excluded and only those referring to what is deemed to be universally present in any society or nearly so, such as water, were included. Blasi et al. found that 30 items in the Swadesh List tend to be represented (or not represented) with certain sounds among 4,298 languages, represented by 6,447 wordlists. Table 4.1 shows their results. Positive symbol means the sound tends to be used for representing the corresponding concept and negative symbol means that the sound tends not to be. Out of 100 lexical items of the Swadesh list, 40 are present in all of the 6,447 wordlists (where the corresponding lexical item is available), but the remaining 60 are only present in 328 wordlists. The items that belong to those 60 are marked with an asterisk. The remaining 70 items where they did not find any phonosemantic correlation are shown in table 4.2. The vast size of the database employed in their study makes it difficult to argue against the solid fact that certain sounds are disproportionately frequent or rare in words for certain meanings. This disproportional distribution alone is already a counter-argument against the complete arbitrariness of the sign. If the sound-meaning association was completely arbitrary, i. e. random, then all sounds would be evenly distributed to every single meaning. Many of the phonosemantic correlations we see in Blasi et al.’s results match with previous experimental studies on sound symbolism. Rounded body parts (‘knee’ and ‘breast’) have rounded vowels as positive symbols, which corresponds to the previous mentioned perceptual association between round shape and rounded vowels. The positive correlation between ‘small’ and /i/ reflects the vowel-size correspondence shown by numerous studies since Sapir [9]. The association between ‘small’ and /tʃ/ also concurs with LaPolla’s [12] experiment which showed

16 Table 4.1: Concepts that show significant phonosemantic correlations, found by Blasi et al.

Concept Positive symbol Negative symbol Ash* u — Bite* k — Bone k j Breasts u m a h r Dog s t Drink — a Ear k — Eye — a Fish a — Full p b — Hear ŋ — Horn k r — I ɲ u p b t s r l Knee o u p k q — Leaf b p l — Name i o p Nose u n a One t n — Red r — Round* r — Sand* s — Skin — m n Small* i tʃ — Star z — Stone t — Tongue e ɛ l u k Tooth — b m Water — t We n p l s You (sg) — u o p t d q s r l that acute consonants are related to smallness. Moreover, many correlations are fairly intuitively explicable, even if they have not been experimentally demonstrated. ‘Nose’ may have /n/ as the positive symbol because when articulating /n/ (and other nasal sounds), the nasal cavity vibrates. Previously, Urban [27], based on a “genealogically balanced sample” of 111 languages, has shown that morphemes that stand for ‘nose’ tend to have nasal sounds (he additionally found that morphemes for ‘lips’ tend to have bilabial consonants). The positive relationship between ‘tongue’ and /e, ɛ/ could be explained by the fact that when we stick out our tongue, we go “eh”, similar to when we articulate /e/ or /

17 Table 4.2: Concepts where Blasi et al. did not find any significant phonosemantic correlations

Present in all blood, come, die, drink, ear, of the 6,447 eye, fire, hand, liver, louse, wordlists (where mountain, new, night, path, available) person, see, tree, two Only present in all, bark, belly, big, bird, 328 wordlists black, burn, claw, cloud, cold, (where available) dry, earth, eat, egg, feather, flesh, fly, foot, give, good, grease, green, head, heart, hot, kill, know, lie, long, man, many, moon, mouth, neck, not, rain, root, say, seed, sit, sleep, smoke, stand, swim, tail, that, this, walk, what, white, who, woman, yellow

ɛ/. ‘Fish’ may tend to be represented with /a/ because most fishes keep their mouth wide open, its shape being similar to the human mouth articulating /a/. As much as Blasi et al.’s study provides us valuable insight on lexical iconicity, a similar typological study based on a different wordlist could also be helpful. This is partly because the Swadesh List has a disadvantage considering the purpose of analyzing linguistic iconicity. Swadesh deliberately excluded from his list lexical terms that he deemed to have “sound imitative tendencies” [26, p. 126]. For example, he excluded ‘to blow (air)’ from his list because he judged that words meaning ‘to blow’ tend to have labial consonants and/or sibilants, which are iconically associated with the oral action of blowing air. This was because this iconic tendency could lead to the false impression that two genealogically unrelated words are cognates. While this is a valid criterion helpful for Swadesh’s original purpose, it directly works against the purpose of research on linguistic iconicity. Thus, investigating a different wordlist containing different lexical terms, including those excluded by Swadesh, would further broaden our view on lexical iconicity.

18 4.4 Diachronic iconicity

We have seen that many lexical items prefer sounds that iconically match their meanings and disprefer sounds that contradict their meanings. But how did these correlations emerge? For example, if many morphemes for ‘nose’ have a nasal sound, what is this the result of? Is it the case that the ancient languages, such as Proto-Indo-European, created words for ‘nose’ with nasal sounds out of thin air and modern languages have preserved them to our date? Saussure [2], aside from admitting that onomatopoeic items are not fully arbitrary, notes that some non-onomatopoeic lexical items may show some non-arbitrary relationship between sound and meaning. One of his examples is French fouet /fwɛ/ ‘whip’, whose pronunciation may sound like a whip (presumably resembling the whooshing sound a whiplash makes). But he points out that fouet is etymologically not onomatopoeic and not originally motivated by sound symbolism: it is derived from Latin fāgus ‘beech’. He distinguishes such cases from “real onomatopoeic items” (which have onomatopoeic origins) and claims that a posteriori resemblance like that of fouet is merely a “chance result of phonetic evolution.” (p. 102) Saussure’s presupposition that a posteriori resemblance between sound and meaning is necessarily a “chance result” remains questionable. In fact, they may be not due to chance, but due to a general tendency towards iconicity. Jespersen [28] makes a long list of words that include the vowel i and mean something related to smallness (we have previously seen several experimental studies showing the perceptual correlation between high front vowels and small size). He does not claim that all of the words he lists have necessarily had i or meant something small from the beginning. Instead, he believes that many cases of this phonosemantic correlation may be what Saussure referred to as “chance result of phonetic evolution”, i. e. that they may have ended up having i and meaning smallness or something small through sound and semantic change. But he argues that the results are not due to pure chance but to the general preference to preserve iconic words and discard non-iconic ones:

[…] I am not speaking of the origin or etymology of the words enumerated: I do not say that they have from their very first taken their origin from a desire to express small things symbolically. It is true that some of the words mentioned have

19 arisen in that way,––many of our i-words are astonishingly recent––but for many others it is well-known that the vowel i is only a recent development, the words having had some other vowel in former times. What I maintain, then, is simply that there is some association between sound and sense in these cases, however it may have taken its origin, and however late this connexion may be (exactly as I think that we must recognize secondary echoisms). But I am firmly convinced that the fact that a word meaning little or little thing contains the sound [i], has in many, or in most, cases been strongly influential in gaining popular favour for it; the sound has been an inducement to choose and to prefer that particular word, and to drop out of use other words for the same notion, which were not so favoured. In other words, sound-symbolism makes some words more fit to survive and gives them a considerable strength in their struggle for existence. […] In this way languages become richer and richer in symbolic words. I do not believe in a golden first age in which everything in language was expressive and had its definite significative value, but rather in a slow progressive tendency towards fuller and easier and more adequate expressions (also emotionally more adequate expressions) —and in this movement the increasing number of sound-symbolisms forms to my mind a not inconsiderable element. (p. 288)

This hypothesis seems persuasive when we look at the typological evidence that many non- onomatopoeic words show iconic characteristics. For example, previous research shows that nasal sounds are frequent in words meaning ‘nose’ in the world’s languages. If we see a nasal sound in a lexical item for ‘nose’ in a certain language, there are two possible causes behind it. The first possible cause is that this lexical item was originally a mimetic word expressing something related to the nose with one or more nasal sounds ever since it was created and then later changed into the non-ideophonic basic word for ‘nose’. The second possible cause is that the lexical item formerly did not have a nasal sound and/or did not mean anything related to ‘nose’, but later ended up having a nasal sound and the meaning ‘nose’ through language change, and it survived, thanks to its later-gained iconicity. If thousands of words for ‘nose’ show a statistic preference for nasal sounds, this preference

20 cannot be due to the first cause alone, because basic words like those for ‘nose’ are seldom newly created, rarely borrowed, and usually very old [29]. This means that they usually go through a very long process of sound change, which makes it relatively difficult to preserve their nasal sound which they may have had from their mimetic origin. Thus, the other mechanism, as proposed by Jespersen, must be at work. Johansson and Carling [30] demonstrate that the deictic iconicity in Indo-European languages is not directly inherited from Proto-Indo-European but is a result of constant reconstructions throughout history. As previously mentioned, proximal terms (this) tends to have vowels of higher F2 than distal terms (that) do. This is not because there was –in Jespersen’s words –a “golden first age” of iconicity where every word in every language was iconic, including deictic terms, and modern deictic terms are remnants of those ancient iconicity. The iconicity of the contemporary languages reflects continuous language changes that repeatedly strive for iconicity. For example, the Portuguese proximal term est–, medial term ess–, and distal term aquel– are iconic in the sense that the proximal and medial terms’ vowel /e/ has a high F2 whereas the distal term has both /e/ and /a/, the latter having a low F2. But this vocalic iconicity is not directly inherited from Latin: Latin’s proximal, medial, and distal terms were ist–, ips–, and ill–, all having only /i/ as the vowel. Poruguese aquel–results from the combination of the expletive adverb *accu and ill–. By analogy, the expletive adverb was also added to proximal and medial terms in medieval and early modern Portuguese (resulting in aquest–and aquess– ), but these forms did not survive in modern Portuguese; only aquel–did [31] (p. 159). This is arguably because aqu–added only to the distal term would produce a deictic schema that is more iconic. Johansson and Carling show that similar reconstructions motivated by iconicity occur in a statistically significant amount of Indo-European languages. This shows that words are not only born iconic but continuously evolve to be so. By this logic, there is no need to trace the proto-forms of contemporary lexical items to show that contemporary lexical items bear iconic tendencies. For example, we do not have to gather reconstructed words for ‘nose’ from proto-languages to show that words for ‘nose’ tend to have nasals: gathering contemporary words for ‘nose’ would suffice, because our languages are no less iconic than our ancestors’ languages were. Linguistic iconicity is not a fossilized heritage, it

21 is a living force. In the present study, therefore, I will focus on the iconicity of the contemporary lexical items and not that of the “ancient” lexical items.

22 Chapter 5 Methodology

As already mentioned, the purpose of the present study is to examine if certain sounds occur abnormally frequently or infrequently in lexical items that refer to certain meanings, similar to the study of Blasi et al. My study, however, will use a different wordlist, a different classification of sounds, and a different calculation of the frequency of each class.

5.1 Wordlist

Instead of the Swadesh List, I have chosen to use a wordlist called Leipzig-Jakarta List [29]. This list consists of 100 basic terms, 62 of them overlapping with the Swadesh List items, although some overlapping items differ in details (hand in Swadesh List v. arm/hand in LJ List). The benefits of using a list that partly overlaps with the Swadesh List are twofold. The overlapping terms will allow us to cross-check Blasi et al.’s findings, while the terms not included in the Swadesh List may provide us new information about lexical iconicity. The 100 terms were empirically selected out of a sample of 41 languages based on four factors: Resistance to borrowability. Out of thousands of lexical items of the 41 languages, the researchers of the languages judged which are (likely to be) loanwords and which are not. And the lexical concepts that were less likely to be loanwords were given more credit to be selected as part of the wordlist. Non-analysability. The LJ List has sought to exclude meanings that are likely to be represented by compounds rather than individual morphemes. Hence, there are who? and what? in the list but not other interrogative pronouns, such as where? or why? which, according to Tadmor, is likely to be polymorphemic (as e.g. what-place? for-what?). Universality. The LJ List needed to create a list of concepts that are found in all (or most of) human societies. Thus, animals that are found in all human societies, such as ‘dog’, ‘fish’,

23 ‘fly’, ‘louse’, and ‘ant’ are up on the list, but not the animals that are (or were) absent in some societies, such as ‘cow’, ‘pig’, or ‘cat’. Stability. Lexical items that are older were given more score than younger lexical items. The usage of this list thus minimizes the possibility that a group of morphemes in my data that represent a given lexical meaning shows a tendency to have a certain sound because they are etymologically related. The list’s non-analysability is also compatible with my database, which only contains morphemes and excludes polymorphemic words. Lastly, its universality assures that most of the sample languages, which are typologically diverse, will not lack a morpheme for a given meaning because a certain society is unaware of that meaning. The four factors are gradual, not absolute. The LJ List’s non-borrowability does not mean that there are no loan-morphemes in my data: the Korean morpheme /jʌn/ ‘smoke’ is borrowed from Chinese /jɛnH/, for example. (I have not excluded loanwords because it is impossible for me to detect all the loanwords, especially within the lesser-studied languages, and it is incon- sistent to pick out loanwords only from the languages I have diachronic knowledge of.) Addi- tionally, of the 66 languages studied, not all have a non-analysable morpheme for every single meaning of the LJ List: in some cases, a language has only analyzable compounds for a given meaning. The LJ List is shown in table 5.1, in alphabetical order.

24 Table 5.1: The LJ List.

1sg pronoun 2sg pronoun 3sg pronoun ant arm/hand

ash back big bird to bite

bitter black blood to blow bone

breast to burn (intr.) to carry child (kin term) to come

to crush/grind to cry/weep to do/make dog to drink

ear to eat egg eye to fall

far fire fish flesh/meat fly

to give to go good hair hard

to hear heavy to hide to hit/beat horn

house in knee to know to laugh

leaf leg/foot liver long louse

mouth name navel neck new

night nose not old one

rain red root rope to run

salt sand to say to see shade/shadow

skin/hide small smoke soil to stand

star stone/rock to suck sweet tail

take_to thick thigh this to tie

tongue tooth water what? who?

wide wind wing wood yesterday

5.2 Sample Languages

In order to avoid the risk of morphemes being phonetically similar because they are from the same origin, it was necessary that I select languages from different language families. I made the list of 66 largest language families (based on the number of speakers), based on the data from the 20th Edition of Ethnologue [32], excluding pidgins, creoles, mixed languages, unclassified languages, sign languages, and the constructed language Esperanto. Each language

25 isolate was counted as a language family with a single language member. I have selected the largest language of each language family (based on the number of native speakers), i. e.:

Indo-European (the largest language family) > Spanish (the largest Indo-European language) Sino-Tibetan (the 2nd largest language family) > Mandarin (the largest Sino-Tibetan language) …

For the first ten languages (Spanish to Vietnamese), I have judged their largest status from Ethnologue’s page “Summary by Language Size” (http://www.ethnologue.com/statistics/size, consulted on 25/02/2017). For the remaining 56 languages, I have accessed every language page of each 56 language family on the Ethnologue website to see which one is the largest, except for Yoruba, which I judged to be the largest language of the Niger-Congo family based on Parkvall [33]. This is because Niger-Congo has 1,539 languages, and it would be very time-consuming for me to consult 1,539 language pages. Whenever the Ethnologue page marked the distinction between L1 population and L2 population, I made the choice based on the L1 population number. When there wasn’t any clear notation of L1 and L2 distinction, I relied on the largest number that refers to the number of speakers. Why pick the largest language from each family? In order to systematically select one language per family, I could have used any other method. For example, I could have chosen the language whose name alphabetically comes first or the language that has the longest name. But since larger languages usually have more resources available, selecting the largest language makes the data easier to collect and more credible. Moreover, relying on a factor such as native speaker population, which is largely irrelevant to the nature of the language per se, avoids tilting the group of languages towards a specific linguistic characteristic. Having a systematic criterion –arbitrary, but systematic–also helps to avoid the risk of cherry-picking languages for the sake of proving my hypotheses. Note that the number of native speakers is as it is written in the 20th Edition of Ethnologue, not based on any other source (except for, as I have mentioned, the case of Yoruba). For exam-

26 ple, Woods Cree is noted as having 35,000 speakers on Ethnologue, whereas Greensmith [34] estimates the population to be no more than 8,000. Moreover, some may argue that English has more native speakers than Spanish has. But since the goal of this thesis is not to find out the ex- act number of native speakers and I only need to systematically select one language per family, I have relied on the population data as shown on Ethnologue, as it serves its goal, imperfect it may be. The classification of languages may also be controversial, for the distinction between a dialect and a language is inherently sociocultural and fluid. Again, I have relied on Ethnologue to judge whether a given language is really a single language or not, because it reflects a consensus among many specialists. For example, Chuanqiandian Cluster Miao is, as its name suggests, a cluster of dialects rather than a single language, but I have listed it as a single language because Ethnologue classifies it as such. The list of languages is shown in table 5.2.

Table 5.2: The 66 sample languages.

Family Language [ISO 639-3]

Indo-European Spanish [spa] Sino-Tibetan Mandarin Chinese [cmn] Niger-Congo Yoruba [yor] Afro-Asiatic Egyptian Spoken Arabic [arz] Austronesian Javanese [jav] Dravidian Telugu [tel] Turkic Turkish [tur] Japonic Japanese [jpn] Austroasiatic Vietnamese [vie] Tai-Kadai Thai [tha] Koreanic Korean [kor] Nilo-Saharan Dholuo [luo] Uralic Hungarian [hun]

cont.

27 Table 5.2 cont.

Family Language [ISO 639-3]

Hmong-Mien Chuanqiandian Cluster Miao [cqd] Quechuan South Bolivian Quechua [quh] Mongolic Peripheral Mongolian [mvf] N. Caucasian Kabardian [kbd] Mayan K’iche’ [quc] Tupian Paraguayan Guarani [gug] Kartvelian Georgian [kat] Trans-New Guinea Enga [enq] Uto-Aztecan Eastern Huasteca Nahuatl [nhe] Aymaran Central Aymara [ayr] Otomanguean Meqzuital Otomi [ote] Maipurean Wayuu [guc] Basque Basque [eus] Chibchan Ngäbere [gym] Totonacan Highland Totonac [tos] Khoe-Kwadi Khoekhoe [naq] West Papuan Galela [gbi] Mapudungu Mapudungun [arn] Tarascan Western Highland Purepecha [pua] Algic Woods Cree [cwd] Eyak-Athabaskan Navajo [nav] Mixe-Zoquean Highland Popoluca [poi] Sepik Ambulas [abt] Misumalpan Mískito [miq] Jivaroan Shuar [jiv] Chocoan Northern Emberá [emp]

cont.

28 Table 5.2 cont.

Family Language [ISO 639-3]

Torricelli Bukiyip [ape] Kx’a Northwestern !Kung [vaj] Eskimo-Aleut Greenlandic [kal] Burushaski Burushaski [bsk] East Bird’s Head-Sentani Sentani [set] South Bougainville Terei [buo] Cariban Macushi [mbc] Ramu-Lower Sepik Ap Ma [kbx] Paezan Páez [pbb] Matacoan Wichí Lhamtés Vejoz [wlv] Sandawe Sandawe [sad] Tungusic Xibe [sjo] Guaykuruan Toba [tob] Ticuna Ticuna [tca] Jean Kaingang [kgp] Australian Pitjantjatjara [pjt] Guajiboan Guahibo [guh] Panoan Shipibo-Conibo [shp] Yanomaman Yanomamö [guu] Tucanoan Tucano [tuo] Warao Warao [wba] Barbacoan Awa-Cuaiquer [kwi] Maybrat Mai Brat [ayz] Sálivan Piaroa [pid] Border Amanab [amn] Muskogean Choctaw [cho]

cont.

29 Figure 5.1: Distribution map of sample languages. Longitude and latitude of each language were retrieved from Glottolog 3.0 [1]. Map created with Microsoft Excel 2016.

Table 5.2 cont.

Family Language [ISO 639-3]

Iroquoian Cherokee [chr]

One disadvantage of this sampling is that areal distribution wasn’t taken into consideration. Languages that are areally close to each other may influence each other’s lexicon. Among my sample languages, several languages are areally clustered in Mexico and Papua New Guinea. In other areas, however, the languages are fairly well spread out from each other. The distribution is shown in figure 5.1.

5.3 Database

The database used for the present analysis is different from the database used in Blasi et al.’s study, not only in size but also in terms of what is included in and excluded from the database.

30 In order to avoid selection bias and unequivocally reflect the basic vocabulary of the sample languages, I needed to set certain principles for gathering the database. Unlike the database of Blasi et al. which lists words that signify the Swadesh List meanings, my database only lists morphemes for the LJ List meanings, excluding grammatical affixes. This is because grammatical affixes are arguably irrelevant to the iconic relation between the lexical stems and the meanings they designate. For example, in the Chinese word shéng-zi ‘rope’, the nominal suffix–zi is largely irrelevant to the specific meaning of ‘rope’. In order to examine the phono-semantic correlations of the lexical item for ‘rope’, it seems reasonable to exclude what mainly serves morphological purposes, in this case the suffix–zi. Polymorphemic words are not listed, because I hypothesized that an analyzable polymorphemic word cannot be deemed iconic in the same way as a non-analyzable morpheme is. For instance, the Thai word for ‘fly (insect)’, malaeng-wan, is a compound consisting of malaeng ‘bug’ and wan ‘day’, literally ‘day bug’. Since this word refers to the meaning ‘fly’ indirectly through the combination two morphemes that do not refer to ‘fly’, it seemed to me inadequate to put such word in parallel with other morphemes that directly refer to ‘fly’. If a language has no corresponding morpheme for a meaning, I have left the slot empty. In some cases, a language has more than one morphemes to express one meaning. In those cases, I have listed up to three morphemes for each meaning per language. A definition of a morpheme is the meaningful recurrent unit that can be compounded with, but cannot be divided into, other meaningful recurrent units. The English segment /ð–/ recurs in this, that, these, those, and the and is related to the meaning of definiteness, but is not a morpheme since it is not compounded with other morphemes but only paired with meaningless segments like /–ɪs/ or /–æt/. This criterion is not categorical but rather gradient. Morphemes sometimes merge with non- recurrent segments as well, such as in cranberry, where cran– does not recur anywhere else, whence the term “cranberry morphemes.” But these cases are exceptions rather than the rule, and most of the time, the abovementioned criterion serves as a consistent tool for determining what a morpheme is. Morphemes that were compounds historically but do not consist of recurrent meaningful

31 morphemes synchronically are counted as single morphemes. In Spanish /sopl/ ‘to blow’ (< Lat. sub ‘under’ + fl ‘to blow’), /pl/ is not a recurrent morpheme, and although the prefix / so–/ ‘under’ is meaningful and recurrent (/so + kaβ/ ‘to dig under,’ /so + fɾei/ ‘to fry lightly’), its semantic relationship to /sopl/ is no longer evident. /sopl/, therefore, is better viewed as a morpheme. I have only selected morphemes that are used in the neutral context, avoiding literary or vulgar words. For example, for the Mandarin Chinese morpheme for ‘eye’, I have selected yǎn 眼 rather than mù 目 because yǎn occurs in the generic word for ‘eye’ yǎn-jing 眼睛 (lit. ‘eye-eyeball’) whereas mù does not. The same criterion applies for personal pronouns. I have only listed person pronouns that can be used between close friends of the same age and gender in a neutral context and spoken language (as far as my knowledge of each language permits me to make this distinction). In some cases, such pronoun does not exist: there is no such Japanese 2nd person pronoun or Korean 3rd person pronoun. In these cases, I left the slots empty. Some languages lexically distinguish human and animal body parts. I have not listed terms that exclusively refer to non-human body parts, except for inherently non-human body parts (‘wing’, ‘tail’, ‘horn’, and ‘hide’). Furthermore, I have set specifications for some of the LJ List terms, shown in table 5.3.

32 Table 5.3: Specifications for some of the LJ List terms.

Term Specification

Back The body part, not the direction Breast What primarily refers to the female organ, rather than what primarily refers to the gender-neutral body part (i. e. ‘chest’). Child (kin term) Can be substituted by ‘son’ and ‘daughter’ only if a language has no commonly used corresponding gender-neutral term. In The affix or adposition meaning “in a location or a container.” To know To have the knowledge, not to be acquainted with a person. Old Both the oldness of an object and of a person. Thick The cylindrical thickness (as in “thick stick”) and the surface thickness (as in “thick surface”), but not density (as in “thick hair”). 1sg, 2sg, and 3sg pronouns Unbound morphemes in nominative, absolutive, or emphatic case. If there is no unbound morpheme as the pronoun, the nominative or absolutive bound morpheme is selected.

When a language only has a corresponding morpheme with a different morphological category, I have listed it. For example, the Yoruba has no verbal morpheme (to my knowledge) that means ‘to laugh’ but only the noun /rĩH/ ‘laughter,’ which I listed as ‘to laugh’. Due to data availability, I could not find all the morphemes available. There are 66 languages and 100 meanings, thus 6,600 “slots” to fill (or leave empty when a language does not have an appropriate morpheme for a meaning). Out of the 6,600 slots, 112 were left empty due to lack of data. The database is thus approximately 98% complete.

33 5.4 Transcription

Generally speaking, most of what is lexical was transcribed. Long vowels and geminates are transcribed into two segments (VV and CC) and not with the long symbol “ː” (Vː and Cː). This is because longer segments are arguably cognitively more salient than shorter ones, which suggests that more credit should be given to their presence. Tones are uniformly represented as L (low), M (mid), H (high), R (rising), and F (falling). Very low tones and very high tones are transcribed as LL and HH, respectively. Contour tones may be R, F, or one of these two combined with L, M, or H (e. g. HR as high rising tone). Stresses were not transcribed, even though stress makes minimal pair distinctions in some languages (e. g. Japanese). This is because the stress was not transcribed in many sources available. Diacritics that serve no phonemic purpose are generally not transcribed (such as post-nasal vowel nasalizations). Coarticulated segments were not transcribed as a character with a diacritic (e. g. /pʷ/) but as two normal-sized characters (e. g. /pw/).

5.5 Classification

In order to calculate the occurrence of a certain class of sounds in morphemes referring to a certain meaning, it is necessary to first determine what is meant by class. A class refers to certain segments grouped together based on their articulatory features. For consonants, I have classified them according to the articulator and the manner of articulation. For example, labials and nasals are two classes. For vowels, I have classified them based on their height, backness, and roundedness: for instance, high vowels, front vowels, and unrounded vowels are three classes. I have included classes that are common throughout the world’s languages. For example, I have included approximants as a separate class, but not clicks, because approximants are common whereas clicks are relatively uncommon. I have determined which classes are “common” in the following way. PHOIBLE Online [35] shows that each of the following phonemes is present in the inventory in at least 70% of the 2155 sample languages. We can thus say that these are relatively “common”

34 phonemes.

Table 5.4: Phonemes that exist in the inventory of at least 70% of 2155 languages.

Phoneme Number of languages where the phoneme exists/The total number of sample languages (per- centage)

m 2053/2155 (95%) k 2016/2155 (94%) i 1998/2155 (93%) a 1961/2155 (91%) j 1901/2155 (88%) u 1873/2155 (87%) p 1873/2155 (87%) w 1812/2155 (84%) n 1742/2155 (81%) s 1663/2155 (77%) t 1604/2155 (74%) b 1533/2155 (71%)

If a phoneme occurs in the inventory of at least 70% of the world’s languages, then by corollary, it means that the articulatory classes where that phoneme belongs are present in at least 70% of the world’s languages. For example, /m/ is present in 95% of the languages. We can thus say that the classes of labials and nasals exist in at least 95% of the languages. By this logic, we can make a list of classes that exist in at least 70% of the languages, shown in table 5.5. (The glides were treated as consonants.)

35 Table 5.5: Phonemes and classes that exist in more than 70% of the languages.

Phoneme Class

m labial, nasal k dorsal, stop i high, front, unrounded a low, front, unrounded j dorsal, approximant u high, back, rounded p labial, stop w labial, dorsal, approximant n coronal, nasal s coronal, fricative t coronal, stop b labial, stop

If we sum up the classes listed in Table 7, then we have a list of classes that are phonemically present in the inventory of at least 70% of the world’s languages and therefore may safely be referred to as common classes.

Table 5.6: Common classes

Consonant–Articulator labials, coronals, dorsals Consonant–Manner stops, fricatives, nasals, approximants Vowel–Height high, low Vowel–Backness front, back Vowel–Roundedness rounded, unrounded

There are thus 13 classes in total that we can safely assume to be “common”. This thesis seeks to examine how frequent the segments belonging to each of these classes are in every set of morphemes for a specific meaning.

36 The specific IPA symbols that belongs to each class are listed in table 5.7. Again, glides were treated as consonants. Each character stands for every occurrence of that character, including within aspirated, ejective, affricates, coarticulated sounds, etc. For example, /p/ stands for not only the independent occurrence of /p/, but also its occurrence in /ph/, /p’/, /pf/, /kp/, and so on. Secondary articulations are not distinguished from full-fledged segments. For example, the secondary articulation /w/ in /pw/ is not distinguished from /w/ as an independent articulation and is counted as a segment.

37 Table 5.7: IPA characters of each class.

Classes IPA characters

Labials p, b, m, ʙ, ɸ, β, ɱ, ⱱ, f, v, ʋ, Articulator ʘ, ɓ, w, ʍ, ɥ Coronals t, d, n, r, ɾ, θ, ð, s, z, ʃ, ʒ, ɬ, ɮ, Consonants ɹ, l, ʈ, ɖ, ɳ, ɽ, ʂ, ʐ, ɻ, ɭ, ǀ, !, ǂ, ǁ, ɗ, ɕ, ʑ, ɺ, ɧ Dorsals c, ɟ, ɲ, ç, ʝ, j, ʎ, k, g, ŋ, x, ɣ, ɰ, ʟ, q, ɢ, ɴ, ʀ, χ, ʁ, ʄ, ɠ, ʛ, ɧ, w, ʍ, ɥ Stops p, b, t, d, ʈ, ɖ, c, ɟ, k, g, q, ɢ, ʔ, ʡ Manner Fricatives ɸ, β, f, v, θ, ð, s, z, ʃ, ʒ, ʂ, ʐ, ç, ʝ, x, ɣ, χ, ʁ, ħ, ʕ, ʜ, ʢ, ɕ, ʑ, ɧ, ɬ, ɮ, ʍ Nasals m, ɱ, n, ɳ, ɲ, ŋ, ɴ Approximants ʋ, ɹ, ɻ, j, ɰ, l, ɭ, ʎ, ʟ, w, ɥ

High i, y, ɨ, ʉ, ɯ, u Height Low a, ɶ, ɑ, ɒ Front i, y, e, ø, ɛ, œ, æ, a, ɶ Vowels Backness Back ɯ, u, ɤ, o, ʌ, ɔ, ɑ, ɒ Rounded y, ʉ, u, ʏ, ʊ, ø, ɵ, o, œ, ɞ, ɔ, ɶ, Roundedness ɒ Unrounded i, ɨ, ɯ, ɪ, e, ɘ, ɤ, ə, ɛ, ɜ, ʌ, æ, ɐ, a, ɑ

38 5.6 Proportion Calculation

The proportion of a class X in morphemes for given meaning in a given language is calcu- lated as follows:

∑ |m| |x in mi| i=1 |segments of mi| |m|

where m = morpheme that stands for the meaning, segment = one IPA symbol used for transcription, x = a segment that belongs to the class X, and | | = cardinality. Each diacritic is counted as one segment. Aspirated consonants, affricates, nasal vowels, long vowels, geminates, diphthongs, and coarticulated segments are counted as two segments. Affricates are counted as two segments consisting of a stop and a fricative. For example, the affricate /ts/ is counted as a stop /t/ plus a fricative /s/. Each tone initial (H, M, L, R, and F) is a single segment. As an illustration, suppose that we wish to find the proportion of labials in morphemes for ‘ant’ in language X, and this language has three morphemes for ‘ant’:

1. pa (2 segments, 1 labial)

2. mata (4 segments, 1 labial)

3. ta (2 segments, 0 labial)

Then the proportion of labials in morphemes for ‘ant’ in language X would be

1/2 + 1/4 + 0/2 = 0.25 3

This calculation method is effective because it controls for morpheme length and the overall number of morphemes each language has and also reflects the number of sounds within a morpheme. The longer a morpheme is, the more likely that it contains a given sound by chance, so a longer morpheme should be given less credit for containing that sound. Similarly, a language can have more than one morpheme that expresses a given meaning, whereas another language has only one, which could lead to a disproportionately large influence from the language with

39 more morphemes if we do not control for the number of morphemes. Moreover, we can also give more credit to morphemes which contain more than one instance of a given sound, which arguably reflects the degree of iconicity better than a binary distinction between those that have a given sound and those that don’t.

5.7 Statistical Analysis of Correlations

With the proportions, I conducted a two-tailed Z-test for every class, the sample being the proportions of a sound for a given meaning, and the population being all the proportions of a sound for all 100 meanings. As an illustration, suppose that 0.1 is the average proportion of nasals in all the existing morphemes for 100 meanings, and 0.05 the standard deviation. And suppose that 65 out of 66 languages have a morpheme for ‘nose’, and the average proportion of nasals in all 65 morphemes for ‘nose’ is 0.12. In this case, the Z-test for nasals in ‘nose’ is conducted as:

0.12 − 0.1 z = 0.05 , p≈0.00126 √65 The Z-tests were followed by multiple comparisons. There are 100 meanings and 24 features, thus 2400 Z-tests. I conducted two Benjamini-Hochberg procedures [36], one at False Discovery Rate (FDR) of 10% and another at FDR of 5%. The phonosemantic correlations represented by the p-values at different FDRs are interpreted as shown in table 5.8.

Table 5.8: The phonosemantic correlations represented by the p-values at different FDRs.

10% ≤ FDR 5% ≤ FDR < 10% FDR < 5%

Correlation Insignificant Weak Strong

40 Chapter 6 Results and Discussion

6.1 Positive and Negative Correlations

Table 6.1 shows the lexical meanings that show positive or negative correlations with any of the 24 classes. “Positively correlated” terms are terms that have significantly high proportions of the class on the left, and “negatively correlated” terms are terms that have significantly low proportions. Weak (5% ≤ FDR < 10%) correlations are in gray text and strong (FDR < 5%) correlations in black text. Out of 100 terms, 36 of them show at least one correlation.

Table 6.1: Positive and negative correlations.

Classes Positively correlated Negatively correlated

to blow Labials to laugh thigh

to suck Coronals navel to go small

to suck Dorsals – in navel

hard 3sg thigh Stops not to tie tongue thick

cont.

41 Table 6.1 cont.

Classes Positively correlated Negatively correlated

salt to suck Fricatives in to blow small

1sg 2sg to eat in to blow Nasals nose leaf not skin/hide this breast name

to cry/weep Approximants breast wind

ant to blow not High fire wide navel nose

navel to suck Low not breast knee neck

cont.

42 Table 6.1 cont.

Classes Positively correlated Negatively correlated

in to blow to say knee this Front navel what? neck who? to suck flesh/meat

navel neck to blow Back breast – knee to suck water

breast knee navel Rounded this neck to blow to suck

1sg to blow in breast to say knee Unrounded this navel what? neck who? to suck

43 6.2 Comparison with Blasi et al.

It is difficult to directly compare Blasi et al.’s results with mine. They have analyzed individual segments, whereas I have analyzed articulatory classes. It would be nevertheless worthwhile to do a simple comparison. Table 6.2 compares the cross-checks between Blasi et al.’s results and mine. By cross-check, I mean that at least one phoneme Blasi et al. have found to be correlated with a meaning belongs to at least one class I have found to be correlated with the same meaning. For example, the positive symbol /u/ for ‘breast’ found by Blasi et al. is cross-checked by the class “rounded” found in my analysis because /u/ belongs to the class of rounded vowels. The results that are cross-checked by the two studies are in bold text. The positive and negative weak correlations found in my results are in gray text. Note that some “same” lexical terms differ in detail (such as ‘hair’, which in the Swadesh List only refers to human hair, but to both human and non-human hair in the LJ List). Lexical items that are not present in all the 6,447 wordlists but only in 328 wordlists in Blasi et al.’s ASJP database are noted with an asterisk.

Table 6.2: Comparison between the results of the two typological studies

Items Positive (Blasi Positive (Joo) Negative Negative (Joo) et al.) (Blasi et al.)

ash* u – – –

big* – – – –

bird* – – – –

to bite* k – – –

black* – – – –

blood – – – –

bone k – j –

rounded unrounded breast u m nasals a h r low back approximant

44 cont.

Items Positive (Blasi Positive (Joo) Negative Negative (Joo) et al.) (Blasi et al.) to burn* (intr) – – – – to come – – – – dog s – t – to drink – – a – ear k – – – earth* (soil) – – – – to eat* – nasals – – egg* – – – – eye – – a – fire – high – – fish a – – – flesh* (/meat) – front – – foot* (/leg) – – – – to give* – – – – good* – – – – hair* – – – – hand (/arm) – – – – to hear ŋ – – – horn k r – – –

front rounded knee o u p k q – unrounded back low to know* – – – – leaf b p l – – nasals liver – – – –

45 cont.

Items Positive (Blasi Positive (Joo) Negative Negative (Joo) et al.) (Blasi et al.) long* – – – – louse – – – – mouth* – – – – name i nasals o p stops

front rounded neck* – – unrounded back low new – – – – night – – – –

nasals nose u n a – high

nasals stops not* – – low high one t n – – – rain* – – – – red* r – – – root* – – – – sand* s – – –

front to say* – – – unrounded to see – – – – skin (/hide) – – m n nasals

coronals small* i tʃ – – fricatives smoke* – – – –

46 cont.

Items Positive (Blasi Positive (Joo) Negative Negative (Joo) et al.) (Blasi et al.)

to stand* – – – –

star z – – –

stone (/rock) t – – –

tail* – – – –

nasals this* – front – rounded unrounded

tongue e ɛ l – u k stops

tooth – – b m –

water – back t –

front what?* – – – unrounded

front who?* – – – unrounded

nasals 1sg pronoun ɲ u p b t s r l – unrounded

2sg pronoun – nasals u o p t d q s r l

There remains the question why some of my results are not present in Blasi et al.’s study and vice versa. Even for the corresponding results, some of them are only partially corresponding: as for the nasality of 1sg pronouns, it is unclear why among all the nasals, only /ɲ/ is present in Blasi et al.’s results, while Gordon [23] and Nichols and Peterson [24] confirm that /n/ and /m/ are also prominent.

47 6.3 Possible Motivations

6.3.1 Mouth

Terms that are related to oral actions or oral organs show correlations to sounds whose articulation involve related oral actions. ‘To blow’ is strongly positively correlated with labials and weakly so with fricatives and rounded vowels, perhaps because when we blow air out of our mouth, we protrude our lips, burst them open, and hiss out air. As previously mentioned, this has been predicted by Swadesh [26], who observed that ‘to blow’ tends to have labial and sibilant consonants. On the other hand, ‘to laugh’, which usually involves wide mouth open- ing, is strongly negatively correlated with labials. ‘To suck’ is strongly positively correlated with fricatives and weakly so with rounded vowels, perhaps because sucking things causes oral friction and lip protrusion. ‘Nose’ is strongly positively related to nasals, arguably because we vibrate our nasal cavity when articulating nasal sounds. This correlation is also attested both in Urban [27] and Blasi et al. [25].

6.3.2 Shape

Rounded vowels and are positively correlated with round shapes, be it circular (‘navel’), spherical (‘breast’, ‘knee’), or cylindrical (‘neck’). As previously introduced to the reader, experimental studies have already shown the perceptual correlation between rounded vowels and round shape. Blasi et al. also discovered /o/ and /u/ to be positive symbols for ‘knee’ and /u/ for ‘breast.’ They did not find any positive or negative symbol for ‘neck’, however.

6.3.3 Size

Coronals and fricatives are weakly positively correlated with ‘small’. This is cross-checked by the correlation between ‘small’ and /tʃ/ (a coronal fricative) found by Blasi et al.. As mentioned in Section 2, LaPolla’s experiment [12] shows that participants associate acute consonants with smallness. Acute consonants largely overlap with coronal consonants. Furthermore, Klink’s experiment [15] showed that /v/ sounds smaller than /g/ does, which gives us the pos-

48 sibility that fricatives in general sound smaller than non-fricatives, although this needs further detailed verification. We see that neither ‘small’ or ‘big’ is correlated with vowels in any way, unlike what previous experimental studies have shown [9, 10, 37]. On the other hand, this concurs with Bauer’s typological study [38], which confirms that vowel-size correspondence is not significantly reflected in evaluative morphology (diminutives and augmentatives). Although Blasi et al. did find /i/ to be a positive symbol for ‘small’, they did not find any symbol to be related to ‘big’. This suggests that what is shown in experimental studies may not necessarily be visible in typological tendencies.

6.3.4 Texture

Stops seem to be related to firmness. They are strongly positively correlated with ‘hard’ and ‘to tie’. Stops are a very “firm” articulation, since they involve constricting the articulator before bursting out air. Stops may also be related to thickness. They are weakly positively correlated with ‘thick’ and strongly so with ‘thigh’, which is a thick body part. On the other hand, approximants seem to be related to fluidity. ‘Wind’ (fluid movement of air) and ‘to cry/weep’ (fluid movement of tears) are both strongly positively correlated with approximants. This may be because when articulating an approximant, the air (fluidly) escapes from the thin gap of the articulators.

6.3.5 Deixis

‘This’ and ‘in’, both denoting proximity, are both strongly positively correlated front and unrounded vowels. This corresponds to the previous typological studies on deictic terms [22, 20, 21], which suggested that vowels with higher F2 are common in proximal terms, whereas the opposite is more common in distal terms. The fact that ‘this’ and ‘in’ also strongly prefer nasals suggests that nasals are also related to proximity. Nasals are strongly positively correlated with 1sg and 2sg pronouns. Pronominal nasality, as already introduced, is attested in several previous studies. 3sg pronouns strongly disprefer

49 stops, which, to my knowledge, has not been attested before.

6.3.6 Verbal signs

Some correspondences found seem to match with non-linguistic verbal signs. ‘Not’ may be positively correlated with nasals because we commonly use nasal sounds to respond to yes-no questions (mm for yes, m-m! for no). This may be because we use minimal articulatory effort to express simple affirmation or negation, and the velum is by default in a lowered state, as Dingemanse et al. [39] have pointed out. The strong preference of ‘to eat’ for nasals may be also found in the fact that the vocal expressions of pleasant eating are remarkably similar cross-linguistically, sharing nasal sounds: En- glish yum yum, Korean nyam-nyam, French miam-miam, Spanish ñam ñam, Italian gnam gnam, Russian nyam nyam, Vietnamese măm măm, Thai màm màm, Indonesian yumyum, Swedish namnam, Finnish nam-nam, and so on. Anonymous (p. c.) has suggested that the reason may be that when the mouth is busy chewing and swallowing food, the nose is the only organ left to vocally express satiety. This also explains why nasals are weakly positively correlated with ‘breast’: it is usually the first thing a baby ever “eats”. The two interrogative words in the list (‘what?’ and ‘who?’) are both strongly positively correlated with front and unrounded vowels. This largely concurs with Dingemanse et al.’s analysis [39], which shows that interjections for other-initiated repair (the huh? expression, used by the speaker when s/he has not heard clearly what the other speaker had said) in ten languages all have only unrounded vowels, mostly front and/or low. Dingemanse et al. also cited 21 other similar interjections in different languages, and the interjections all have only unrounded vowels. (The y used for transcribing the Tai/Lue interjection is not a high front rounded vowel /y/, but a high mid/back unrounded vowel /ɨ ɯ/. cf.[40, 41]). On the other hand, ‘what?’ and ‘who?’ do not show a preference for low vowels. This suggests that front and unrounded vowels are frequently used in interrogative expressions in general, not just in the interjections for other-initiated repair.

50 Chapter 7 Conclusion

This thesis is provides additional supporting evidence for the importance of iconicity in the spoken language lexicon. Linguistic signs, far from being entirely arbitrary, show varying degrees of iconicity, and iconic mappings between forms and meaning are not limited to the margins of language, instead occurring in basic lexical items. The work presented here demon- strates, in line with previous research, that iconicity can facilitate statistical regularities in the lexicons of the world’s languages. The present study provides further insights on lexical iconicity that were not visible in Blasi et al.’s study. It was able to confirm that the lexical terms that Swadesh [26] has predicted to be highly iconic and excluded from the Swadesh List do bear phonosemantic tendencies. In fact, all the terms that Swadesh excluded due to iconicity and are present in the LJ List have shown preference for or against certain classes: ‘to blow’ (for fricatives, labials, and high/back/rounded vowels and against nasals and front/unrounded vowels), ‘to cry/weep’ (for approximants), and ‘to laugh’ (against labials). Furthermore, the analysis based on articulatory classes, rather than individual segments (which Blasi et al.’s analysis was based on), provides us a different perspective on phonosemantic tendencies. This study can also motivate future experimental studies. Although some correlations have already been experimentally demonstrated (such as magnitude symbolism), other correlations (e. g. between stops and thickness) are yet to experimentally studied. If these typological patterns concur with experimental results, we will be able to confirm that they are psychologically real. In this thesis, I found that for one hundred lexical terms across 66 languages, 36 out of 100 lexical terms showed evidence of a systematic mapping between articulatory classes and meanings. The recognition that iconic associations between words and meanings are so common changes how we view the linguistic sign: humans are not unbiased machines taking in formless inputs and spitting out behaviour–rather we have strong biases that make certain mappings seem more natural, which can have implications for language learning and use. The work presented

51 here suggests that the mapping between language and cognition is not idiosyncratic, but follows patterns that can better help us understand human nature.

52 References

[1] Harald Hammarström, Robert Forkel, and Martin Haspelmath. Glottolog 3.0. http:// glottolog.org, 2017.

[2] Ferdinand de Saussure. Cours de Linguistique Générale. Payot, Paris, 1916.

[3] Paola Pietrandrea. Iconicity and Arbitrariness in Italian Sign Language. Sign Language Studies, 2(3):296–321, 2002.

[4] Edward S. Klima and Ursula Bellugi. The Signs of Language. Harvard University Press, Cambridge, Massachusetts; London, England, 1979.

[5] Sarah F. Taub. Language from the Body: Iconicity and Metaphor in American Sign Lan- guage. Cambridge University Press, Cambridge, 2001.

[6] Fritz Strack, Leonard L Martin, and Sabine Stepper. Inhibiting and facilitating conditions of the human smile: A nonobtrusive test of the facial feedback hypothesis. Journal of personality and social psychology, 54(5):768, 1988.

[7] Mark Dingemanse, Will Schuerman, Eva Reinisch, Sylvia Tufvesson, and Holger Mit- terer. What sound symbolism can and cannot do: Testing the iconicity of ideophones from five languages. Language, 92(2):e117–e133, 2016.

[8] Hoai Dung Nguyen, Thi Minh Nguyet Bui, Thi Thanh Tam Nguyen, and Tran Thuy Van Đoan. Animal onomatopoeias in English and Vietnamese. Tạp chí Khoa học - Trường ĐH Quy Nhơn, 10(3):107–114, 2016.

[9] Edward Sapir. A study in phonetic symbolism. Journal of Experimental Psychology, 12(132):225–239, 1929.

[10] Kazuko Shinohara and Shigeto Kawahara. A cross-linguistic study of sound symbolism: The images of size. In Annual Meeting of the Berkeley Linguistics Society, volume 36, pages 396–410, 2010.

53 [11] John J. Ohala. The frequency codes underlies the sound symbolic use of voice pitch. In John J. Ohala, Leanne Hinton, and Johanna Nichols, editors, Sound Symbolism, pages 325–347. Cambridge University Press, Cambridge; New York; Melbourne; Madrid; Cape Town; Singapore; São Paulo, 1994.

[12] Randy J. Lapolla. An experimental investigation into phonetic symbolism as it relates to Mandarin Chinese. In Leanne Hinton, Johanna Nichols, and John J Ohala, editors, Sound Symbolism, volume 130, page 147. Cambridge: Cambridge University Press, Cambridge; New York; Melbourne; Madrid; Cape Town; Singapore; São Paulo, 1994.

[13] Douglas H Whalen and Andrea G Levitt. The universality of intrinsic F 0 of vowels. Journal of phonetics, 23(3):349–366, 1995.

[14] Roman Jakobson. Six Leçons Sur Le Son et Le Sens. Éditions de Minuit, Paris, 1976.

[15] Richard R Klink. Creating brand names with meaning: The use of sound symbolism. Marketing Letters, 11(1):5–20, 2000.

[16] Wolfgang Köhler. Gestalt Psychology: An Introduction to New Concepts in Modern Psychology. Liveright, New York, 1947.

[17] Mathilde Fort, Alexander Martin, and Sharon Peperkamp. Consonants are More Impor- tant than Vowels in the Bouba-kiki Effect. Language and Speech, 58(2):247–266, 2015.

[18] Ozge Ozturk, Madelaine Krehm, and Athena Vouloumanos. Sound symbolism in infancy: Evidence for sound-shape cross-modal correspondences in 4-month-olds. Journal of Ex- perimental Child Psychology, 2012.

[19] Russell S. Tomlin. Basic Word Order (RLE Linguistics B: Grammar): Functional Prin- ciples, volume 13. Routledge, London; Sydney; Wolfeboro, New Hampshire, 2014.

[20] Christine Tanz. Sound symbolism in words relating to proximity and distance. Language and speech, 14(3):266–276, 1971.

[21] Nancy L Woodworth. Sound symbolism in proximal and distal forms. Linguistics, 29(2): 273–300, 1991.

54 [22] Niklas Johansson and Jordan Zlatev. Motivations for Sound Symbolism in Spatial Deixis: A Typological Study of 101 Languages. The Public Journal of Semiotics, 5(1):3–20, 2013.

[23] Matthew J Gordon. The phonological composition of personal pronouns: Implications for genetic hypotheses. In Annual Meeting of the Berkeley Linguistics Society, volume 21, pages 117–128, 1995.

[24] Johanna Nichols and David A Peterson. The Amerind personal pronouns. Language, pages 336–371, 1996.

[25] Damián E Blasi, Søren Wichmann, Harald Hammarström, Peter F Stadler, and Morten H Christiansen. Sound-meaning association biases evidenced across thousands of languages. Proceedings of the National Academy of Sciences of the United States of America, 113(39):10818–23, 2016.

[26] Morris Swadesh. Towards Greater Accuracy in Lexicostatistic Dating. International Journal of American Linguistics, 21(2):121–137, 1955.

[27] Matthias Urban. Conventional sound symbolism in terms for organs of speech: A cross- linguistic study. Folia Linguistica, 45(1), 2011.

[28] Otto Jespersen. Symbolic Value of the Vowel i. In Selected Writings of Otto Jespersen (Routledge Revivals), pages 515–535. Routledge, London; Tokyo, [2010] 1921.

[29] Uri Tadmor. Loanwords in the world’s languages: Findings and results. In Martin Haspel- math and Uri Tadmor, editors, Loanwords in the World’s Languages: A Comparative Handbook, pages 55–75. The Hague: De Gruyter Mouton, 2009.

[30] Niklas Johansson and Gerd Carling. The de-iconization and rebuilding of iconicity in spatial deixis: An Indo-European case study. Acta Linguistica Hafniensia, 47(1):4–32, 2015.

[31] Milton M Azevedo. Portuguese: A Linguistic Introduction. Cambridge University Press, 2005.

55 [32] Gary F. Simons and Charles D. Fennig. Ethnologue: Languages of the World, Twentieth Edition. SIL International, Dallas, Texas, sil intern edition, 2017.

[33] Mikael Parkvall. Världens 100 största språk 2010. http://www.ne.se/spr%C3%A5k/ v%C3%A4rldens-100-st%C3%B6rsta-spr%C3%A5k-2010, 2010.

[34] Jennifer M Greensmith. Phonological Variants in Pukatawagan Woods Cree. PhD thesis, University of Manitoba, 1985.

[35] Steven Moran, Daniel McCloy, and Richard Wright. PHOIBLE Online. Max Planck Institute for Evolutionary Anthropology, Leipzig, 2014.

[36] Yoav Benjamini and Yosef Hochberg. Controlling the false discovery rate: A practical and powerful approach to multiple testing. Journal of the royal statistical society. Series B (Methodological), pages 289–300, 1995.

[37] Klemens Knoeferle, Jixing Li, Emanuela Maggioni, and Charles Spence. What drives sound symbolism? Different acoustic cues underlie sound-size and sound-shape mappings. Scientific Reports, 7(1):5562, 2017.

[38] Laurie Bauer. No phonetic iconicity in evaluative morphology. Studia linguistica, 50(2): 189–206, 1996.

[39] Mark Dingemanse, Francisco Torreira, and N. J. Enfield. Is ”Huh?” a universal word? Conversational infrastructure and the convergent evolution of linguistic items. PLoS ONE, 8(11):1–10, 2013.

[40] Mary Rosamond Haas. Thai-English Student’s Dictionary. Stanford University Press, Stanford, CA, 1964.

[41] Michael Moerman. The preference for self-correction in a Tai conversational corpus. Language, pages 872–882, 1977.

[42] Eugenio Martínez-Celdrán, Ana Fernández-Planas, and Josefina Carrera-Sabaté. Illus- trations of the IPA: Castilian Spanish. 33(2):255–259, 2003.

56 [43] San Duanmu. The Phonology of Standard Chinese. Oxford University Press, Oxford; New York, second edition, 2007.

[44] Kayode J. Fakinlede. English-Yoruba, Yoruba-English: Modern Practical Dictionary. Hippocrene Books, New York, 2003.

[45] Virginia Stevens and Maurice B Salib. A Pocket Dictionary of Spoken Arabic of Cairo. American University in Cairo Press, Cairo ; New York, new rev. edition, 2004.

[46] E Sukanda, V Sukanda-Tessier, and A Suprijanto. Parlons Javanais: Langues, Dia- logues, Lexiques. Editions L’Harmattan, Paris, 2016.

[47] Stuart Robson and Singgih Wibisono. SEAlang Library Javanese Dictionary. http:// sealang.net/java/dictionary.htm, 2002.

[48] Deena Bossé and Olivier Bossé. Parlons Télougou. Editions L’Harmattan, Paris, 1994.

[49] Lavanya Collooru. Telugu-English/English-Telugu Dictionary & Phrasebook. Hip- pocrene Books; Bilingual edition, New York, 2017.

[50] John Peter Lucius Gwynn. A Telugu-English Dictionary. Oxford University Press, Delhi, 1991.

[51] Bh. Krishnamurti. Telugu. In Sanford B. Steever, editor, The Dravidian Languages, pages 265–311. Routledge, London ; New York, 2015.

[52] A. D. Alderson and Fahir İz. The Oxford English-Turkish Dictionary. Oxford University Press, Oxford, second edition, 1978.

[53] Karl Zimmer and Orhan Orgun. Turkish. Journal of the International Phonetic Associa- tion, 22(1-2):43–45, 1992.

[54] Laurence Labrune. The Phonology of Japanese. Oxford University Press, Oxford ; New York, 2012.

[55] Laurence C Thompson. A Vietnamese Reference Grammar, volume 13. University of Hawai’i Press, Honolulu, 1987.

57 [56] Shoichi Iwasaki and Inkapiromu Puriyā Horie. A Reference Grammar of Thai. Cambridge University Press, New York, 2005.

[57] Ben G. Blount and Elise Padgug-Blount. LUO - ENGLISH DICTIONARY, with Notes on Luo Grammar. Institute of African Studies, University of Nairobi, Kenya.

[58] Alfred Malandra. English-Lwoo Dictionary. Ed. Nigrizia, Bologna, 1957.

[59] Thomas P Gorman. A Glossary in English, Kiswahili, Kikuyu and Dholuo. Cassell, London, 1972.

[60] László Országh and Tamás Magay. A Concise English-Hungarian Dictionary. Oxford University Press & Akadémiai Kiadó, Oxford ; New York ; Budapest, 14th edition, 1990.

[61] Péter Siptár and Miklós Törkenczy. The Phonology of Hungarian. Oxford University Press, Oxford ; New York, 2000.

[62] 兴贵 and 亚东. 代: 言. 贵州民族出版社, 贵阳, 2004.

[63] 熊玉有. 教程. 云南大学出版社, 昆明, 2014.

[64] 石莉芸 and 李云兵. 366 会. 社会科学文献出版社, 北京, 2016.

[65] Teofilo Laime Ajacopa. Diccionario Bilingüe Iskay Simipi Yuyayk’ancha, Quechua- Castellano, Castellano-Quechua. La Paz, 2007.

[66] Joaqúin. Herrero and Federico. Sánchez de Lozada. Diccionario quechua-castellano, castellano-quechua : para hispanohablantes que estudian quechua. Instituto de Idiomas, Padres de Maryknoll, Cochabamba, 1974.

[67] J L Urioste and S J Herrero. Gramática y Vocabulario de La Lengua Quechua. Editorial Canata, La Paz ; Cochabamba, 1955.

[68] Borjigin Sechenbaatar. The Chakhar Dialect of Mongol: A Morphological Descrip- tion. Number 243 in Suomalais-ugrilaisen Seuran toimituksia. The Finno-Ugrian Society, Helsinki, 2003.

58 [69] Jan-Olof Svantesson. The Phonology of Mongolian. The Phonology of the World’s Lan- guages. Oxford University Press, New York, 2005.

[70] Amjad M Jaimoukha and Michel Malherbe. Parlons Tcherkesse: Dialecte Kabarde. Editions L’Harmattan, Paris, 2009.

[71] Buba Macikovich Kardanov and Aslan Tirovich Bichoev. Russko-kabardinsko- cherkesskij slovar’. Gos. izd-vo inostrannyh i nacional’nyh slovarej, Moskva, 1955.

[72] M. L. Apazhev and N. A. Bagov. Kabardinsko-russkij slovar’. Ripol Klassik, Moskva, 2013.

[73] Amjad Jaimoukha. Kabardian Phonetics. jaimoukha.synthasite.com/resources/Kabar- dian Phonetics.doc, 2009.

[74] Abraham Garćia Hernández, Santiago Yac Sam, and David Henne Pontious. Diccionario Quiche–Español. Instituto Linguistico de Verano, Guatemala, 1980.

[75] Allen J Christenson. K’iche’ - English Dictionary and Guide to Pronunciation of the K’iche’-Maya Alphabet. Brigham Young University, 1993.

[76] Antonio. Guasch and Diego. Ortiz. Diccionario castellano-guarani guarańi-castellano : sintactico, fraseologico, ideologico. Centro de Estudios Paraguayos ”Antonio Guasch”, Ascuncion, 2008.

[77] Tʻamar. Gvarjalaże, I Gvarjalaże, Tʻamar. Gvarjalaże, and I Gvarjalaże. English- Georgian Dictionary = Inglisur-Kʻartʻuli Lekʻsikoni. Sakartvelos matsne, Tbilisi, 2003.

[78] Nicholas Awde. Georgian-English/English-Georgian Dictionary & Phrasebook (Hip- pocrene Dictionary & Phrasebook). Hippocrene Books, New York, 2010.

[79] Manuel de. Lucca D. and Manuel de. Lucca D. Diccionario aymara-castellano, castellano-aymara. Comisión de Alfabetización y Literatura en Aymara, La Paz, 1983.

[80] Juan Carvajal Carvajal, Arturo Hernández Sallés, Nelly Ramos Pizarro, Carlos Cárcamo Luna, and Jacqueline Ticona Rojas. Diccionario Ilustrado Aymara, Español, Inglés, volume 3. Pehuén Editores Limitada, Santiago, 2001.

59 [81] Luis Hernández-Cruz, Moisés Victoria Torquemada, and Donaldo Sinclair Crawford. Diccionario Del Hñähñu (Otoḿi) Del Valle Del Mezquital, Estado de Hidalgo. Instituto Lingüístico de Verano, A.C., México, D.F., segunda edición, electrónica edition, 2010.

[82] Ethel Emily Wallis. Diccionario Castellano-Otomí, Otomí-Castellano. Ed. del patrimo- nio indígena del Valle del Mezquital y del Instituto Lingǘistico de Verano, Itzmiquilpan, 1956.

[83] David M. Captain and Linda B. Captain. Diccionario Básico Ilustrado : Wayuu- naiki-Espanol Español-Wayuunaiki. Edit. Fundación para el Desarrollo de los Pueblos Marginados, Bogotá, D. C., primera edición edition, 2005.

[84] Gorka Aulestia, Linda White, Gorka Aulesta, Linda White, Gorka Aulestia, Linda White, Gorka Aulesta, and Linda White. English-Basque Dictionary. University of Nevada Press, Reno, Nevada, 1990.

[85] Ephraim S Alphonse. Guaymí Grammar and Dictionary: With Some Ethnological Notes. Number 162 in Smithsonian Institution Bureau of American Ethnology Bulletin. Bureau of American Ethnology, Washington, 1956.

[86] Herman P. Aschmann. Castellano-Totonaco, Totonaco-Castellano: Dialecto de La Sierra. Instituto Lingǘistico de Verano, México, D.F., segunda edición edition, 1983.

[87] Wilfrid H G Haacke and Eliphas Eiseb. A Khoekhoegowab Dictionary with an English- Khoekhoegowab Index. Gamsberg Macmillan, Windhoek, 2002.

[88] Howard Shelden. Galela phonemes and stress. Workpapers in Indonesian Languages and Cultures, 7:77–88, 1989.

[89] Lućia Golluscio, Adriana Fraguas, and Fresia Mellico. Mapudungun vocabulary. In Martin Haspelmath and Uri Tadmor, editors, World Loanword Database. Max Planck Institute for Evolutionary Anthropology, Leipzig, 2009.

[90] Alberto Medina Pérez and Jesús Alveano Hernández. Vocabulario Español-p’urhépecha, p’urhépecha-Español. Plaza y Valdés, México, D.F., 2000.

60 [91] Claudine Chamoreau. Parlons Purepecha: Une Langue Du Mexique. Editions L’Harmattan, Paris, 2003.

[92] Robert Joseph Castel and David Westfall. Castel’s English-Cree Dictionary and Memoirs of the Elders. Brandon Univ. Northern Teacher Education Program, Brandon, Manitoba, 2001.

[93] Robert W Young and William Morgan. The Navajo Language: A Grammar and Collo- quial Dictionary, volume 3. University of New Mexico Press, Albuquerque, 1980.

[94] Lynda J de Jong Boudreault. A Grammar of Sierra Popoluca (Soteapanec, a Mixe- Zoquean Language). Dissertation, The University of Texas at Austin, 2009.

[95] John Kundama, Adéru Sapayé, and Patricia R. Wilson. Ambulas Dictionary. 2006.

[96] H. Berckenhagen. Pocket Dictionary. Miskito-English Spanish and English-Miskito- Spanish. Moravian Mission in Bluefields, Nic., Bethlehem, PA., 1906.

[97] C. R. Heath and W. G. Marx. Diccionario Miskito-Español Español-Miskito. Imprenta Calderon, Tegucigalpa, 1953.

[98] M. B. Borman. Listas Comparativas de Palabras En Diez Idiomas Autóctonos Ecuato- rianos. Instituto Lingǘistico de Verano, Quito, 1991.

[99] Instituto Bilingüe Intercultural Shuar. Chicham Nekatai : Diccionario Comprensivo, Castellano-Shuar. Sistema de Educación Radiofónica Bicultural Shuar, Sucúa, 1988.

[100] Solomon I Sara. A Tri-Lingual Dictionary of Emberá-Spanish-English, volume 38. Spot- light Poets, München, Lincom.

[101] Christa König and Bernd Heine. A Concise Dictionary of Northwestern! Xun, volume 21. R. Köppe, Köln, 2008.

[102] Oqaasileriffik and Inerisaavik. Greenlandic English Dictionary. https://oqaasileriffik.gl/ resources/about-dictionaries/greenlandic-english-dictionary/, 2017.

61 [103] Hermann Berger. Das Yasin-Burushaski (Werchikwar), volume 3. Otto Harrassowitz Verlag, Wiesbaden, 1974.

[104] Etienne Tiffou. Parlons Bourouchaski: État Présent Sur La Culture et La Langue Des Bourouchos. Editions L’Harmattan, Paris, 1999.

[105] H. K. J. Cowan. Grammar of the Sentani Language. Martinus Nijhoff, ’s-Gravenhage, 1965.

[106] Donald Laycock. A Dictionary of Buin: A Language of Bougainville. Pacific Linguistics, Canberra, 2003.

[107] Emanuele. Amodio, Vicente Pira, and Patricio Sarmento. Silveira. Língua Makuxi : Makusi Maimu : guias para a aprendizagem e dicionário da língua Makuxi. Diocese de Roraima, Roraima, 1996.

[108] Marianna C Slocum, Florence L Gerdel, and Elisabeth. Seibert. Diccionario páez- español, español-páez. Editorial Townsend, Lomalinda, Meta, República de Colombia, 1983.

[109] Jairo Arias, Loly Colderón, Yesena Rincón Jiménez, Anańia Piñacue, Pedro Maŕin Silva, K. David Harrison, and Jeremy Fahringer. Nasa yuwe Talking Dictionary. http:// www.talkingdictionary.org/paez, 2012.

[110] T Rojas. La Lengua Paez, Una Visión de Su Gramática. Editorial Ministerio de Cultura, Bogotá, 1998.

[111] Verónica Nercesian. Wichi Lhomtes: Estudio de La Gramática y La Interacción Fonoloǵia-Morfoloǵia-Sintaxis-Semántica. LINCOM GmbH, München, 2014.

[112] Alejandra Vidal and Verónica Nercesian. Wichí vocabulary. In Martin Haspelmath and Uri Tadmor, editors, World Loanword Database. Max Planck Institute for Evolutionary Anthropology, 2009.

[113] Christopher Ehret. A Dictionary of Sandawe : The Lexicon and Culture of a Khoesan People of Tanzania. Köppe, R, Köln, 2012.

62 [114] Ryohei Kagaya. A Classified Vocabulary of the Sandawe Language. Institute for the Study of Languages and Cultures of Asia and Africa (ILCAA), Tokyo, 1993.

[115] Veronika Zikmundová. Spoken Sibe: Morphology of the Inflected Parts of Speech. Karolinum Press, Charles University in Prague, 2013.

[116] D. O. 朝克. 代口研究. 民族出版社, 北京, 2006.

[117] 张泰鎬. 研究. 云南民族出版社, 昆明, 2008.

[118] Cristina Messineo. Lengua Toba (Guaycurú): Aspectos Gramaticales y Discursivos. Lincom Europa, München, 2003.

[119] Cristina Messineo. Vocabulario toba de Cerrito (Paraguay). http://pueblosoriginar- ios.com/lenguas/toba.php.

[120] Alberto S. Buckwalter, Lois Litwiller de Buckwalter, Lois Litwiller de Buckwalter, and Lois Litwiller de Buckwalter. Vocabulario Toba. 2nd edition, 2013.

[121] Maŕia Emilia Montes Rodŕiguez. Morfosintaxis de La Lengua Tikuna. Centro Colom- biano de Estudios de Lenguas Aborgenes, Universidad de los Andes, Bogot, 2004.

[122] Ursula Gojtéj Gojtéj Wiesemann. Dicionário Kaingang-Português Português-Kaingang. Editora Evangélica Esperança, Curitiba-PR, 2011.

[123] Cliff Goddard. A Learner’s Guide to Pitjantjatjara/Yankunytjatjara. Institute for Abo- riginal Development, Alice Springs, N.T., 1993.

[124] Cliff. Goddard. Pitjantjatjara/Yankunytjatjara to English dictionary. IAD Press, Alice Springs, N.T., 1996.

[125] Victor Kondo and Riena Kondo. Guahibo phonemes. In Viola G. Waterhouse, editor, Phonemic Systems of Colombian Languages, number 14 in Summer Institute of Linguis- tics Publications in Linguistics and Related Fields, pages 89–98. Summer Institute of Linguistics of the University of Oklahoma, Norman, 1967.

63 [126] James Loriot, Erwin Lauriault, and Dwight Day. Diccionario shipibo - castellano. Num- ber 31 in Serie Lingüística Peruana. Ministerio de Educación and Instituto Lingüístico de Verano, Lima, 1993.

[127] Pilar M Valenzuela. Basic Verb Types and Argument Structures in Shipibo-Conibo. M.A. Thesis, University of Oregon, Eugene, 1997.

[128] Pilar M Valenzuela, Luis Márquez Pinedo, and Ian Maddieson. Shipibo. Journal of the International Phonetic Association, 31(2):281–285, 2001.

[129] Jacques Lizot, Hepëwë, Nõhõkuwë, and Tiyetirawë. Diccionario Enciclopédico de La Lengua Yãnomãmi. Vicariato Apostólico de Puerto Ayacucho, Caracas, 2004.

[130] Henri. Ramirez. A Fala Tukano dos ye’pâ-masa. Inspetoria Salesiana Missionária da Amazônia, CEDEM, Manaus, 1997.

[131] Basilio de. Barral, Basilio de. Barral, and Basilio de. Barral. Diccionario Warao- Castellano, Castellano-Warao. Caracas, 1979.

[132] Peter Ponce. Diccionario Español - Warao. http://www.geocities.ws/isla{_}cangrejo{_} language/, October 2017.

[133] Andrés Romero-Figueroa. A Reference Grammar of Warao, volume 6. Lincom Europa, München, 1997.

[134] Timothy Curnow. A Grammar of Awa Pit (Cuaiquer): An Indigenous Language of South- Western Colombia. Ph. D. Thesis, Australian National University, 1997.

[135] August Kafiar, Daniel C Ajamiseba, and Peter J Silzer. Aam Ro Mai Brat = Perbenda- haraan Kata Bahasa Mai Brat = Mai Brat Vocabulary. Number B. 3 in Publikasi khusus bahasa-bahasa daerah. Universitas Cenderawasih and Summer Institute of Linguistics, Jayapura, 1989.

[136] Pedro J Krisologo B. Manual Glotologico del Idioma Wo’tiheh. Centro de Lenguas Indigenas, Instituto de Investigaciones Historicas, Universidad Catolica ”Andres Bello, Caracas, 1976.

64 [137] George Aaron Broadwell. A Choctaw Reference Grammar. Studies in the anthropology of North American Indians. University of Nebraska Press, Lincoln, 2006.

[138] Cyrus Byington. A Dictionary of the Choctaw Language. Number 46 in Bureau of American Ethnology Bulletin. Smithsonian Institution, Washington, 1915.

[139] Brad Montgomery-Anderson. A Reference Grammar of Oklahoma Cherokee. Doctoral Dissertation, University of Kansas, 2008.

[140] Cherokee Nation. Dikaneisdi (Word List). http://www.cherokee.org/About-The-Nation/ Cherokee-Language/Dikaneisdi-Word-List, September 2017.

65 Appendix A The morphemes

Table A.1: Spanish - Mandarin Chinese - Yoruba

Family/ Language Indo-European/ Sino-Tibetan/ Niger-Congo/

[ISO 639-3] Spanish [spa] Mandarin Chinese Yoruba [yor]

[cmn] ant oɾmiga maHRjiLR ɛɛLraL ɛLɛLrũL arm/hand bɾaθo mano piF ʂouLR akpaH ɔwɔH ash θeniθa hweiLR ɛɛHruH back espalda peiHR ɛLjinL big gɾan taHR n̩ HlaH toHbi bird abe paxaɾo njɑuLR eje to bite moɾd jɑuLR bitter amaɾg khuLR koroL black negɾ xeiH duHduH blood sangɾe ɕiɛLR eLdʒeL to blow sopl bente ʈʂhuiH fiHfeH bone weso kuLR ɛɛgũ breast seno ʐuLR ɔmuH ɔjɔ̃L to burn (intr.) aɾd ʐanHR dʒoH to carry ʎeb taiF ru gbɛH child (kin term) ix ɚHR ɔmɔ ɛLwɛ to come ben laiHR waH bɔL dɛH to crush/grind aplast mol sujHR mɤHR fuHnpa lɔ

njɛnLR to cry/weep ʎoɾ khuH sɔkũH cont.

66 Table A.1 cont. to do/make aθ tswoHR ʃɛ dog per kouLR adʒaH to drink beb tom xɤH mu ear oɾexa ɚLR ɛtiH to eat kom ʈʂʐH dʒeũ dʒe õdʒe egg webo tanH ejĩ eye oxo jɛnLR odʒuH ejĩdʒuH to fall ka dɑuLR lwoHR ʃubuH bɔH dʒaH-

waHleL far lexann remot yɛnLR dʒĩL fire fwego xwoLR inaH fish peθ yHR ɛdʒa flesh/meat karne ʐouHR ɛrã fly moska jiŋHR to give d entreg keiLR to go i tɕhyF tsouLR lɔ good bwen xɑuLR rere daHra hair kabeʎo pelo faLR irũ hard dur jiŋF liHlɛ

(s)he/it (3sg) el eʎa eʎo thaH oH to hear oi thiŋH gbɔH heavy pes ʈʂʊŋHR wiHwuHwo

wuHwo to hide eskond twoLR tsaŋHR bo luHgɔ moLkũL to hit/beat golp peg taLR gbaH lu horn asta kweɾno tɕjɑuLR iLwo house kasa tɕjaH wuH faŋHR ilɛH

I (1sg) ɟʝo wɤLR mo in en tsaiF liLR niH cont.

67 Table A.1 cont. knee ɾodiʎa ɕiH orũHkũH eeHkũH to know sab ʈʂʐH mɔL to laugh ɾei ɕɑuHR rĩH leaf oxa jɛF ɛwɛH ɛwɛHko leg/foot pjeɾna pje tɕjɑuLR irɛL eseL liver igado kanH eLdɔL eLdɔLkiH long laɾg ʈʂaŋHR gũL louse pjoxo szH mouth boka tswiLR ɛnu name nombɾe miŋHR oruHkɔ navel ombligo tɕiHR idodo neck kweʎo pɤHR ɔrũL new nweb ɕinH titũ tũtũ night notʃe wanLR jɛHR oLru nose nariθ piHR imuH not no puHR koL old biex maɟʝoɾ antigw lɑuLR tɕiʊLR aLgbaL daLgbaL

gboH one un jiH kã rain ʎubja yLR ɔLdʒɔL edʒi red ɾox hʊŋHR pupa pɔHn root ɾaiθ kənH irĩL ɛgboL eLkɔ̃L rope kweɾda ʂʌŋHR okũL ɔsãH to run kor phɑuLR saHreH suHreH salt sal jɛnHR ijɔL sand aɾena ʂaH jãrĩL ijãĩL to say deθ ʂwoH tɕjɑŋLR so wiH to see be khanF riH cont.

68 Table A.1 cont. shade/shadow sombɾa jinH jiŋLR ibodʒiL iLdʒi

oLdʒiLdʒi skin/hide pjel kweɾo phiHR fuH iLwɔL small pekeɲ tʃik ɕauLR weHweH kɛHrɛH smoke umo jɛnH ɛɛHfĩH soil tjera thuLR ijeLpeL to stand paɾ ʈʂanHR nɔLroH diLdɛ star estreʎa ɕiŋH iLraLwɔL stone/rock pjedra ɾoka ʂʐHR jɛnHR oLkuHta aLpaHta to suck soɾb ɕiH ʂunLR tsaH sweet dulθe tjɛnHR diHdũL alaHdũL tail kola weiLR iLrɛrɛH to take aɣar kox naHR muH thick gɾwes xouHR ipɔ̃ thigh muslo thweiLR itã this est ʈʂɤF ɛLjiH ji15 to tie at khunLR tsaH tongue lengwa ʂɤHR ahɔ̃H tooth djente jaHR ɛhĩH water agwa ʂweiLR omi what? ke ʂʌmHRmɤ who? kjen ʂeiHR wide antʃ khwanH gboLro dʒina wind bjento ajɾe fʌŋH afeHfeH ateHguL wing ala ʈʂhʐHR wood madeɾa muHR paHkoH yesterday aɟʝeɾ tswoHR aLnaH you (2sg) tu niLR o

69 Table A.2: Egyptian Spoken Arabic - Javanese - Telugu

Family/ Language Afɾo-Asiatic/ Austronesian/ Dravidian/ Telugu

[ISO 639-3] Egyptian Spoken Javanese [jav] [tel]

Arabic [aɾz] ant næml ʔiid ʂəmut tʃiima arm/hand diɾææʕ ləŋən tʃejji ash ɾæmææd awu buuɽida back dˤæhɾ gəgəɽ ʋennu big kibiiɾ gəɖe pedda bird tˤeeɾ manuk kiiɾam piʈʈa pakʂi to bite ʕdˤdˤ tʃokot kaɾatʃu kaaθu bitter mʊɾɾ pait tʃeedu black ʔɛswɪd sooda iɽəŋ nalla naluɸu blood dæmm gətih ɾaktam to blow hbb damu ʋiitʃu bone ʕædˤm baluŋ emoka breast sidɾ sˤædɾ ʂuʂu ɽommu to burn (intr.) kaaltsu to carry ʃVl mojju child (kin term) ʕæjjil botʃah pilla pɪllaɽu to come gV təka ʋattsu ɾaa to crush/grind fɾm tˤħn ɽəʒək gəɽuʂ giliŋ ɾaatʃu nuuɾu to cry/weep ʕɛɛtˤ taŋis eeɖtsu to do/make ʕml gawe tseeju dog klb aʂu kokka to drink ʃɾb ombe taagu ear wɪdn kupiŋ tʃeʋi to eat ʔkl kl maŋan tinu egg beedˤ ənɖog guɖɖu cont.

70 Table A.2 cont. eye ʕeen mata kannu to fall wʔʕ tʃəbɭok gigɽig paɽu

guguɽ far bɪʕɛɛd adoh duuɾaŋgaa fire nææɾ gəni nippu manʈa fish sæmæk iwak tʃeepa flesh/meat læħma dagiŋ maansam fly dɪbbææn ɭaɭəɽ iiga to give dd awɛh ittsu to go ɾħ ɭuŋa ʋeɭɭu good kwæjjɪs enak mantʃi baaɣundi hair ʃæʕɾ ɽambut dzuʈʈu hard gææmɪd atoʂ gaʈʈi

(s)he/it (3sg) hʊwwæ hɪjjæ ɖɛwɛke atanu iði aame to hear smʕ kɽuŋu ʋinu heavy tiʡiil abot baɾoʋaina to hide xbb umpət daagu daatʃu to hit/beat xbtˤ dˤɾb kəpruk koʈʈu horn ʔæɾn ʂuŋu kommu house beet omah illu

I (1sg) ʔænæ aku neenu in fɪ iŋ loo knee ɾʊkbæ ɖəŋkuɭ mookaalu to know ʕɾf erti teliju telusu ʋattsu to laugh dħk guju naʋʋu leaf wæɾæʔ goɖoŋ aaku leg/foot ɾɪgl ʂikiɭ kaalu paaðam liver kɪbd kæbɪd ati kaaleejam long tˤæwiil dawa poɽaʋu cont.

71 Table A.2 cont. louse ʔæml tuma peelu mouth buʔʔ tʃaŋkəm nooɾu name ʔism ʒənəŋ peeɾu navel sʊɾɾæ udel wudel boɖɖu neck ɾæʡæbæ guɭu meɽa new gɪdiid aɲaɽ kotta night leel bəŋi ɾaatɾi nose mænæxiiɾ iɽuŋ mukku not mɪʃ dudu oɽa kaadu old ʔædiim tuwa paata one wææħid ʂiʒi oka rain mætˤæɾ udan ʋaɾʂam red ʔæħmæɾ abaŋ eɾupu eɾɾa root gɪzɾ ojot ʋeeɾu rope ħæbl taɭi taaɽu to run gɾ plaju paɾigettu salt mælħ ujah wuppu sand ɾaml paʂiɽ isuka to say ʡl tʃaɭaʈu tʃəɭaʈu omoŋ tʃeppu anu to see ʃf wəɽuh tsuuɽu shade/shadow dˤill bajaŋ niiɖa skin/hide gild kuɭit tsaɾmam small sˤuɣajjaɾ tʃiɭik tʃinna smoke duxxææn kukus poga soil æɾdˤ ləmah neela to stand nilabaɖu nilutʃu star nɪgmæ ɭintaŋ tʃukka taara stone/rock ħægæɾ kaɽaŋ waɖaʂ watu raaji to suck lħs əmpeŋ əntʃit keɲit piiltʃu cont.

72 Table A.2 cont. sweet ħilw ləgi tiipi tail deel buntut tooka to take xd ʒupuk tiisukonu thick tɪxiin kandəɭ dappa thigh fæxd pupu toɽa this dæ dɪ iki idi to tie ɾbtˤ kaʈʈu tongue lɪsææn iɭat naaloka naalɪxa tooth sɪnnæ untu pannu water mæjjæ baɲu niiɭɭu dzalam what? ʔee ʔeeh apa eemiʈi who? miin ʂapa eʋaɾu wide wææsɪʕ amba weɽalpu wind hæwæ aŋin gaali wing gɪnææħ ʂwiwi ɾekka wood xæʃæb kaju kaɾɾa yesterday ʔɪmbææɾɪħ wiŋi ninna you (2sg) ʔɪntæ ʔɪntɪ kowe nuʋʋu miiru

73 Table A.3: Turkish - Japanese - Vietnamese

Family/ Language Turkic/ Turkish Japonic/ Japanese Austroasiatic/ Viet-

[ISO 639-3] [tur] [jpn] namese [vie] ant kaɾɯndʒa ari kiɘnHR arm/hand kol̴ el̴ ɯde te tɐjM ash cyl̴ hai ʈʂɔHR back aɾka se lɯŋM big iɾi byjyc kodʒaman ooki tɔM lɤnHR bird kuʃ tori cimM to bite ɯsɯɾ diʃle hakam kɐnHR ŋwamLʔ bitter adʒɯ cescin niga ɗʕɐŋHR black kaɾa sijah kɯro ɗʕɛnM blood kan tʃi mɐʊHR hwiɘtHR to blow es ɸɯk thoɪLR bone cemic hone sɯɤŋM breast ɟœys tʃitʃi tʃibɯsa vʊH to burn (intr.) jak jan moe ʈʂɐɪHR to carry taʃɯ hakob vakHR xwʌnM

maŋM child (kin term) tʃodʒuk javɾu ko kɔnM to come ɟel var kɯ ɗʕe̞ nHR tɤjHR

laɪLʔ to crush/grind ez œyt tsɯbɯs kɯdag ɛpHR vɐtHR ŋiɘnL to cry/weep aala nak kɐʊkpHR to do/make jap et kɯl̴ sɯ tsɯkɯ lamL dog cœpec it inɯ cɔHR to drink itʃ nom uəŋHR ear kulak mimi taɪM to eat je tabe ɐnM cont.

74 Table A.3 cont. egg jumuɾta tamago ʈʂɯŋM eye ɟœz me mɐtHR to fall dyʃ otʃi rɤɪM far uzak too saM fire ateʃ çi lɯəLR fish bal̴ɯk sakana kaHR flesh/meat et nikɯ thitL fly sinec icikanatl̴ɯ hae ruøɪM to give ver baɯʃl̴a kɯre age cɔM tɐŋLʔ to go ɟit ik ɗˤɪiM good iji ɟyzel i totHR haɪM hair satʃ kami ke tɐʊkpHR loŋM hard katɯ kata kɯŋHR rɐnHR

(s)he/it (3sg) o nɔHR to hear iʃit kik ŋɛM heavy aɯɾ omo nɐŋLʔ to hide sakl̴a sakl̴an kakɯ zɐʊHR to hit/beat vuɾ dœv tʃarp nagɯ tatak ɯt ɗˤɐɲHR horn boj tsɯno sɯŋL house ev hane konut ie ɲaL

I (1sg) ben wataʃi toɪM taM tɤHR in de ni ɤLR taɪʔL

ʈɽɐʊŋmM knee diz hiza ɣoɪHR to know bil ʃi waka biɘtHR to laugh ɟyl waraw kɯɤɪL leaf japrak ha laHR leg/foot badʒak ajak aʃi cʌnM liver kaɾadʒieɾ kimo ɣanM cont.

75 Table A.3 cont. long uzun naga ʒaɪL louse bit ʃirami rʌnLʔ mouth aɯz kɯtʃi momL miɘŋLʔ

mɔmHR name isim ad nam na te̞ nM ʒɐɲM navel ɟœbec heso ronHR neck bojun kɯbi koLR new jeni ataraʃi arata mɤɪHR night gedʒe jorɯ ɗˤe̞ mM toɪHR nose buɾun hana muɪHR not me na xɐʊŋmM old ihtijaɾ jaʃlɯ ɸɯrɯ zaL one biɾ çito motL rain jaamuɾ ame mɯəM red kɯrmɯzɯ kɯzɯl aka ɗˤɔLR root cœk ne re̞ HR rope ip tʃɯna ʒʌɪM to run koʃ haʃi cɐɪLʔ salt tuz ʃio muøɪHR sand kum sɯna katHR to say sœjle i nɔɪHR to see ɟœɾ bak mi thʌɪHR ʈʂʌʊŋmM

ɲinL shade/shadow ɟœlɟe kage ɓɐʊŋmHR skin/hide deɾi dʒilt ten kawa hada ʒaM ɓɪiM small cytʃyc ufak tʃiisa ɲɔLR ɓɛHR smoke duman kemɯri xɔɪHR soil topɾak jeɾ tsɯtʃi ɗˤʌtHR to stand tat ɗˤɯŋHR cont.

76 Table A.3 cont. star jɯl̴dɯz ɟœkdʒismi hoʃi ʂɐʊM stone/rock taʃ kaja kylte iʃi ɗˤaHR to suck em masset sɯ mutHR ɓʊuHR

hutHR sweet tatlɯ ama ŋɔtL tail kujɾuk ʃippo o ɗˤuøɪM to take al̴ moraw kʌmM nɐmHR

zɯHR thick kal̴ɯn cesif atsɯ ʒɐɪM thigh but momo ɗˤuɪL veɪHR this bu ko nɐɪL to tie baal̴a ʃiba ɓuəkL kotL ʈʂɔɪHR tongue dil ʃita lɯɤɪHR tooth diʃ ha rɐŋM water su mizɯ nɯɤkHR what? ne nani ziL who? kim dare aɪM wide ɟeniʃ çiro rʌʊŋmLʔ wind ɾyzɟaɾ jel kaze zɔHR wing kanat tsɯbasa kɐɲHR wood odun tahta ceɾeste ki ɣoʊHR yesterday dyn kinoo you (2sg) sen siz mɐjL

77 Table A.4: Thai - Korean - Dholuo

Family/ Language Tai-Kadai/ Thai Koreanic/ Korean Nilo-Saharan/

[ISO 639-3] [tha] [kor] Dholuo [luo] ant motH gemi bije arm/hand khɛɛnR mʉʉM phal shon bat ash thawF dʑe buru back laŋR dɨŋ tok big jajL kh maŋɔngɔ bird nokH she wɪɲɔ to bite katL mul kajo bitter khomR s maketʃ black damM kama gʌm teŋ blood lʉatF looMhitL phi rɛmɔ to blow phatH phul kʊðɔ bone kraLduukL pjʌ tʃogo breast thanM dʑʌdʑ θuno to burn (intr.) tha waŋ to carry bɛɛkL nal tiŋo child (kin term) dekL ai e mɪhia ɲaθɪ to come maaM o biro to crush/grind biipL kotL atL ɨke tɔjɔ rego to cry/weep hajF ul jwak to do/make thamM ha mandɨl tɪmɔ dog suLnakH gemi gwɔk to drink dʉʉmL maɕhi maðɔ ear huuR gɥi it to eat kinM mʌg tʃamo tʃiemo egg kha2j al tɔŋ eye taaM nun waŋ cont.

78 Table A.4 cont. to fall tokL lonL lwar far klajM mʌl bor fire fajM bul matʃ fish plaaM gogi rɛtʃ flesh/meat nʉaH shal gogi riŋo fly phali lwaŋni to give hajF dʑu mijo to go pajM ga tʃomo ði good diiM dʑo mabɛr hair phomR khonR thʌl jier hard khɛŋR dandan tagtag matɛk

(s)he/it (3sg) khawR thəəM ɛn o

manM to hear jinM dɨd windʒo heavy nakL mugʌb pek to hide sɔɔnF sukH shum pando butno pondo to hit/beat tiiM topL tɕhi teli tʃwado horn khawR pul tuŋ house baanF dʑib ɔt

I (1sg) tʃanR na a an in najM e i knee khawL mulɨb tʃɔŋ to know ruuH al ŋejo to laugh huaRrɔH ush ɲɪɛrɔ leaf bajM iph oboke leg/foot khaaR thawH dali tɪɛlɔ liver tapL gan tʃuɲ long jaawM gil bor louse hawR matL loonM i ɲuogo oɲuogo cont.

79 Table A.4 cont. mouth paakL ib ðɔk name tʃhʉʉF naamM ɲɪŋ navel saLdʉʉM il pel neck khɔɔM mog ŋut tok new majL she maɲien night dʉʉnM bam otieno nose tʃaLmukL kho um not majF an ɔk old kɛɛL nɨlg nalg mapim one nʉŋL hana han atʃiel rain fonR bi kɔθ red dɛɛŋM palgah kwar root raakF huaR puli rope tʃʉakF dʑul tol to run wiŋF dalli rɪngɔ salt klʉaM shogɨm tʃumbi sand saajM mole kwojo to say phuutF bɔɔkL ha ko watʃɔ

klaawL to see henR duuM bo nɛnɔ shade/shadow romF ŋawM gɨnɨl gɨlimdʑa ombre skin/hide phiwR naŋR del pien small lekH dʑak madundo smoke khuanM jʌn iro soil dinM hɨlg lowo loo to stand jʉʉmM shʌ tʃuŋ star daawM bjʌl sulue sulwe stone/rock hinR dol kidi to suck omM duutM pal ɲɔðɔ cont.

80 Table A.4 cont. sweet waanR dal mamɪt ŋar tail haaŋR plaajM koli ip to take jipL tʃapL khuaaH bad kao thick naaR dukʌb gulg pek thigh hʌbʌg em this niF niH i ma to tie phuukL matH muk me tuejo twejo tongue linH hjʌ lep lewni tooth fanM i lak water namH mul pi what? aLrajM mwʌ muʌsh aŋɔ who? khrajM nugu ŋano ŋa wide kwaaŋF nʌlb laatʃ wind lomM balam jamɔ wing piikL buom wood majH namu jɪɛn yesterday waanM ʌdʑe ɲoro you (2sg) thəR nʌ in

81 Table A.5: Hungarian - Chuanqiandian Cluster Miao - South Bolivian Quechua

Family/ Language Uralic/ Hungarian Hmong-Mien/ Quechuan/ South

[ISO 639-3] [hun] Chuanqiandian Bolivian Quechua

Cluster Miao [cqd] [quh] ant hɔnɟɔ ɳtʂouMR sik’imira sisi arm/hand kɔr kɛz teL maki riera ash hɔmu tʂhouHH utʃpha back haat eHH wasa big khɨ4 loF hatun bird mɔdaar noŋLR p’isqu to bite xɔrɔp toLR keuM khani khatʃu p’ata bitter kɛʃɛru aHF tʃaqaq haja hajaq black fɛkɛtɛ tloHF jana blood veer ɳtʂhaŋHH jawar to blow fuuj tʂhuɑHF phuku bone tʃont tshaŋH tuʎu breast mɛll ɛmløø miL ɲuɲu to burn (intr.) eeg tɕiR larwa lawra to carry vis saalliit ɴqaŋH apa child (kin term) ɟɛrɛk ɟɛrmɛk ɲuaLR ususi tʃuri to come jøn tuɑF hamuj to crush/grind zuuz øørøl dɔraal kuta ɲat’u p’alta to cry/weep kiaalt kiɔbaal ordit quɑF waqa to do/make tes tʃinaal keesit uɑH rura ruwa hina dog kuca ɛb tleHH aʎqu to drink isik vɛdɛl houM upjaj ukjaj upij ear fyl ɳtʂeF ninri to eat esik nɑoF mikhuj egg tojaaʃ qeH runtu cont.

82 Table A.5 cont. eye sem n̥ aHH ɲawi to fall poŋHF urma far taavoli tleHF ɲawpa kart fire tyyz teuL nina fish hɔl ɳtʂeL tʃaʎwa flesh/meat huuʃ ɴqaiF ajtʃa fly leeɟ mao21 tʃ’uspi to give ad touH qu qupu to go mɛɟ moL ri good joo ʐoŋH sumaq hair hɔj plouHF kawaʎu phiʎi tʃuk-

tʃa hard kɛmeeɲ teuHH anaq saʎu

(s)he/it (3sg) øø ɔz ɔzt niL paj to hear hɔll n̥ aoHH ujari heavy nɛheez ʃuujoʃ ɲ̥ aŋHH ʎasa to hide rɛjt ɛldug ʐaiM paka to hit/beat yt vɛr ntouM taka maqa matʃ’a horn sarv ɔgɔntʃ koHF waqra house haaz tʂeHH wasi

I (1sg) een koHH ɲuqa in bɔn bɛn ɲaoHF xouHH pi knee teerd houHHtɕouR qunqur to know tud pouHF jatʃa to laugh nɛvɛt kɔtsɔg ʈoM luɑR asi leaf lɛveel mploŋF ʎaqhi raphi leg/foot laab teuH tʃanka tʃaka tʃaki liver maaj ʂaHF k’iwitʃa k’iwtʃan

kukupin cont.

83 Table A.5 cont. long hoʃsuu nteHH sini sajt’u louse tɛtyy toHH usa mouth saaj ɲtɕouF simi name neev mpeH suti navel køldøk nteuM pupu neck ɲɔk kunka new uuj tʂhaHF musuq night eesɔkɔ eej eejjɛl tuta tʃ’isi nose orr ɳtʂuR sinqa not nɛm tʂiH mana old ørɛg veen idooʃ louL qoHF kuraq thantha

mawk’a one ɛɟ iHF huk rain ɛʃøø piroʃ naŋR para red vørøʃ laHF puka root ɟøkeer tɕaŋR saphi rope køteel l̥ uaH lasu to run fut sɔlɔd roxɔn ntouF ʈaŋR phawa salt ʃooʃ ɳtʂeHH katʃi sand homok føvɛɲ ʂuɑHFtsiHH t’iwu aqu to say mond jɛlɛnt haiM ɲi to see laat muhu shade/shadow aarɲeek tluɑHF ʎanthu skin/hide bøør irxa teuH qara small kis kitʃi ɔproo ʑouH jutʃ’uj smoke fyst ɳtʂoM q’usɲi soil tɔlɔj føld luaMF haʎp’a to stand eeg ʂeuHH k’ana ruphatʃi star tʃillɔg n̥ oHFqoHF qujʎur cont.

84 Table A.5 cont. stone/rock køø siklɔ sirt ʐeHF rumi to suck siiv upi tʃ’onqa sweet eedɛʃ qaŋHFʐiHF misk’i tail fɔrok fɔrk tuHH rabo tʃupa to take vɛs muɑHF hap’i thick vɔʃtɔg tuɑHF rakhu thigh tsomb tʃanka this ɛz naHH kaj to tie køt kaeHF kiHH wata tongue ɲɛlv mplaiR qaʎu tooth fog n̥ aHH kiru water viz tleF jaku unu what? mi mit laHFtʂiH ima who? ki toLtuR pi wide seelɛʃ tlaŋHH sakha wind seel tɕuɑ4 wajra wing saarɲ tiM rjera pharpa rapra wood fɔ ntoŋH k’uʎu yesterday tɛgnɔp qajna you (2sg) tɛ kɑoF qam

85 Table A.6: Peripheral Mongolian - Kabardian - K’iche’

Family/ Language Mongolic/ Periph- N Caucasian/ Mayan/ K’iche’

[ISO 639-3] eral Mongolian Kabardian [kbd] [quc]

[mvf] ant ħwmpʔətsʔədʒ sanik arm/hand kar ʔə q’ab ash uns jaʑə tʃax back xœœn txi rix big ix in nim bird ʃʊpʊ qwaləbzw to bite tʃʊʊ xatʃ dzəq katso tijoʔ bitter kaʃʊŋ didʒ didʒ didʒ k’a black xar fʔitsʔə q’eq blood tʃʊs ɮi kik’ to blow qepɕ jepɕ ʃut’ux bone jas qwpɕħə baq breast bʁə to burn (intr.) ʃɪth is nikow to carry ħi child (kin term) ur sabij ral ralk’ual to come ir qəkʔwə petik pe to crush/grind qw qaxbe ʃukle xok’o to cry/weep ʏɪl ʁwəɣ oq’ik to do/make ɕʔ bano dog nɔxœ ħə ts’iʔ to drink ʊʊ ʊntaa jəf kumux uk’ia ear tʃix tħəkʔwmə ʃikin to eat it ʃx waʔ tix egg ontək dʒəditʃə sakmol cont.

86 Table A.6 cont. eye nut nə waq’ʌtʃ to fall pʊʊr dʒəl texw tsak far xɔl ʒiʑə nax fire kal mafʔə q’aq’ fish tʃakəs bdzəʑjəj kʌr flesh/meat max ɮi tiʔx fly badzə amalo to give ok ti jaʔ jaʔo to go jap kʔw ʌbek eʔk good sææŋ fʔi gwapə uts hair us ɕħəts kiex wiʔ hard katʊʊ bidə ko

(s)he/it (3sg) ar areʔ to hear sɔnsɔ zəxəx taʔo ta heavy xunt wəndəʁw ħəɮə al to hide nʊʊ ʁəpɕkʔw awax k’uʔ to hit/beat ɔn jəw nak’baʔ nak’o

tʃ’ajo horn bʑə ruk’aʔ house kɤr unə xa

I (1sg) pii sə in in təəm im pa knee ɮəgwaʑə tʃ’ek to know mɤt ɕʔ retaʔm to laugh ɤnjɤ diħəʃx tseʔnik leaf nɛptʃh tħəmpə ʃoq leg/foot xol ɮə ɮaqwə pʌraʔ liver tħəmɕʔiʁw poror seseb raqan long ʊrth tʃʔiħ roqolik cont.

87 Table A.6 cont. louse poos tsʔe sʌquk’ mouth ʑə tʃiʔ name nɤr tsʔə biʔ navel xiis binʒe muʃuʃ neck pɕə tʃilo new ʃin ɕʔə k’ak’ night son ʒəɕ aq’ab nose pə tsaʔm not qim old xʊʊtʃhəŋ ʑi k’el riʔx one nək zi xun rain wəʃx xʌb red ʊlaaŋ pɮiʑ kiak root qwəps rab k’amal ʃera rope tɤɤs tʃʔapsə k’am to run guii ʒ salt ʃiʁw ats’am sand pʃaħwə sʌnjeb to say ʒiʔ psəɮ bix to see ʊtʃ ɮaʁw krilo shade/shadow nibʑ muʔx skin/hide fə ts’umal ts’um small tʃitʃik pak tsʔikʔw alax smoke ʔwʁwə sib soil ʃɔrɔɔ ɕʔigwɮ ulew to stand pɔs ɕit tak’iʔk star vaʁwə tʃ’imil stone/rock tʃhʊlʊ mivə abax to suck ɕʔəf ts’uʔ cont.

88 Table A.6 cont. sweet ʔəfʔ qiʔ tail suul tʃʔə xeʔ to take ap aw ɕt tʃap thick tʃʊtʃaaŋ ʔwb ʁwm pim sorosik thigh kwə raʔ this ɤn mi riʔ to tie ʊj xulj zəridz px jut’u tongue xɤl bzəgw raq’ tooth ʃut dzə ware water ʊs psi xaʔ what? juu sit xas su who? xɤŋ xət xatʃin wide xan bʁwə lik’ilik wind xii sæljxjaŋ ʑi kiaqiq’ wing damə xik’ wood pħə tʃeʔ yesterday otʃhoktor tʃhoktor diʁwasə iwir you (2sg) tʃhii wə at

89 Table A.7: Paraguayan Guarani - Georgian - Enga

Family/ Language Tupian/ Kartvelian/ Geor- Trans-New Guinea/

[ISO 639-3] Paraguayan gian [kat] Enga [enq]

Guarani [gug] ant akekẽ taxɨi tʃ’iantʃ’wela ɑLɽuHmɑLɽiL arm/hand ʒɨʋa mk’lav xel kiHŋgiL ash kusuɣwe aʋatisoka iLtɑHteL keL-

nɑLiL back kupe lo4mo teɾek- zurg mɑLiRtɑL

waʔ̃ big eɾado ɣwasu didi ɑLndɑHkeL

ŋgwasu jɑRɽeL bird ɣwɨɾa prinvel jɑHkɑL to bite suʔu k’ben n bitter heʔẽmbɨ mts’are kɑLɑL black hũ ʃav puHmbuRtiL blood tuɣuɨ sisxl tɑLeLjoHkoL to blow pɨtu ʃeberw poLoL ɽ bone kã kaŋgwe dzwal kuRɽiL breast pɨtiʔa tapɨpaʔũ titi mk’erd ɑHnduH to burn (intr.) hoʋeɾe kai poɾo- dats’w t

hapɨ to carry ɾu t’areb mɑLnd child (kin term) raʔa bavʃw wɑRneL to come xwaʔĩ mosvl iLp to crush/grind apatuka dapkw iLnd to cry/weep ʒaxeʔo ʒopɨ t’iril

mombe to do/make ʒapo k’eteb piL cont.

90 Table A.7 cont. dog ʒaɣwa dzaɣl jɑHnɑL to drink sɨɾɨku dalev n ear apɨsa nambi q’ur kɑLɽeL to eat ʒepɨʔaʒoko ɲem- tʃ’am n

boapɨte egg a k’wertsx tuRkuH eye tesa ɾesa twal ɽeHŋgeL to fall ʔa mombe moŋgwi datsem mɑLŋg pjɑLk far mombɨɾɨ ʃors ɽoLndɑHtiL fire ʒehapɨ kai ɾata tsetsxl iLtɑHteL fish piɾa tevz piHsɑL flesh/meat soʔo xorts miHndʒuH fly mberu buz poLŋgeRnɑL to give ʒopoi kwapɨtĩ meʔẽ mitsem mɑLiL ndiH to go xo ts’avl p good maɾanɣatu porã k’arg eRpeL hair akãɾagwe aʋa tm iRtiL

tagwe hard atã xatã magar kjeRtoL

(s)he/it (3sg) xaʔe la ʃupe is mbɑLɑL to hear xendu mosmen s heavy pohɨi mdzime keRndɑL to hide moʔã mokaɲɨ damalv

moɲemi to hit/beat xaʋiɾa nupã mboʋu dart’q’m piH iLnd horn tatĩ tuɾu rqa neHŋgeL house oɣapɨ ɾoɣa saxl ɑRndɑL

I (1sg) ʃe me nɑLmbɑL in me pe ʃi ɲɑL cont.

91 Table A.7 cont. knee penarã ɾenɨmɨʔã muxl ɽuHmɑL

tenɨpɨʔã to know kwaa tsnob mɑHs to laugh puka gatsineb leaf toɣwe potol joLkoL leg/foot kupɨ retɨma tetɨma pex moRkoL liver pɨʔa ɣwidzl puHŋgiL long ʋuku grdzel ɽoLndɑHtiL louse kɨ t’il tiHpuH mouth ʒuɾu p’iri kɑRmbuH name ɾeɾa teɾa saxel keHŋgeL navel puruʔã up’e muHmbiLteLŋgeL neck aʒuɾa aʒuɾiɣwa k’iser mɑLŋgeL

atwa new anɣeteɣwa axal eLneHŋgeL night pɨxaɾe ɣame kuHkuH

kuHɑLkɑRmɑL nose apɨŋgwa tĩ tsxwiri ʎɑHɑL not ani ar ndɑLɑL old xaɣuɨno tuʒa dzwel ɑLtiRŋgiL

ɨmaɣwaɾe wɑLmbɑLtɑHeL

ʎɑLŋgɑL one peteĩ maʋa ert meLndɑHiL rain ama ts’wim ɑLiLjuHuH red pɨtã ts’itel root ɨpɨ pesv piHŋgiL rope sã tok’ keRndeL to run mbiɾiki ɲani rben salt ʒukɨtɨ maril ɑLiRpiL cont.

92 Table A.7 cont. sand ɨʋɨkuʔi kwiʃ keHɑL to say e i tkm ɽ to see xeʃa naxw kɑLnd shade/shadow kwaɾaxɨʔã moʔãxa tʃrdili ɽuHmbɑL skin/hide ape apekwe api k’an jɑLnuRŋgiL small tʃuʔi xesaʔi kaɾape p’at’ar jɑLkɑRneL smoke tatatĩ timbo bol keLnɑL soil ɨʋɨ niadagi juHuH to stand hoʋeɾe kai poɾo- dgom

hapɨ star ʒasɨɾata mbɨʒa varsk’vlav mbuHiLiL stone/rock ita kw k’lde kɑLnɑL to suck pɨte sɨɾɨku ts’ov mɑLkɑLpuHs sweet xeʔẽ t’k’bilk teRndeL tail xuɣwai ɾuɣwai k’udi eHteLŋgeL

tuɣwai to take ʒaɣaɾɾa ʒapɨxɨ aɣeb ɲ

ʒaʋoʔo thick xɨpɨʔũ msxwil thigh kwaɾto tetɨma uʋa bardzaqi pɑHiLŋgiL this ko es ndɑHkeL to tie apakwa ʒokwa gamonask’w ɑLndʒ

mosã tongue kũ ena keLkeL tooth ɾãi tãi k’bil neHŋgeL water moak̃ ts’q’al eLndɑHkiL what? xe ra ɑHkiL who? maʋa vin ɑLpiL wide pɨkatu parto cont.

93 Table A.7 cont. wind ɨʋɨtu kari poLoL poLptɑLuH wing pepo prta pɑRpɑL wood ɨʋɨɾa xe iRtɑL yesterday kwexe guʃin kuHɑHkɑL you (2sg) ne peẽ ɾo ʃen eHmbɑL

94 Table A.8: Eastern Huasteca Nahuatl - Central Aymara - Meqzuital Otomi

Family/ Language Uto-Aztecan/ East- Aymaran/ Central Otomanguean/

[ISO 639-3] ern Huasteca Nahu- Aymara [ayr] Meqzuital Otomi

atl [nhe] [ote] ant ɑʃkaneeli k’isimira k’isirmita ʃãxɨ arm/hand mah ampara ʔjæ ash qhiʎa ʔbospi back ati ʃɨtha big weei xatʃ’a dãngi data nda bird tootoo xamatʃ’i ts’ints’ɨ to bite kwah kwaa atʃuqa tsa bitter k’aʎk’u k’ask’a ɲu black tʃ’ijara mboi boi ʔbo blood es wila xi to blow piitsa phusa hwifi bɨ hwiki bone omi tʃ’akha ndoʔjo breast ɲuɲu ʔba ʃiʔba ʔmoʔba to burn (intr.) tɬɑtɬɑ naqha aka lawra zʌ to carry wiikɑ apa tu tuts’i child (kin term) konee phutʃha juqa metsi t’ɨ to come xuta ehe to crush/grind k’ija ɲiti t’una pæt’i kɯni to cry/weep tʃooka xatʃa warari zoni to do/make lura hoki xa ʌt’e dog tʃitʃi anu tsat’yo ʔyo to drink ii uma ʔbæt’i xapi tsi ear nakas xintʃu gu ʃigu to eat kwɑɑ manq’a tsi egg teekkis tʃhuja k’awna mãdo cont.

95 Table A.8 cont. eye iiʃ najra da to fall wetsi liwisi tinku hʌe far xaja ja fire tɬi nina tsibi fish mitʃ tʃawʎa tʃaʎwa hwã flesh/meat naka ajtʃa ngʌ fly saajool tʃhitʃhiʎankha ginʔwe to give mɑkɑ tʃura da uni ʔra to go sara ma pa good kwal aski ʎumpaqa suma hɲo ho hair tson ɲik’uta ʃi stã hard qhulu me

(s)he/it (3sg) j xupa geʔã gehni nuʔã to hear kɑki ist’a ʌde heavy xathi hɲɨ to hide imanta ãgi to hit/beat wiiteki nuwa fant’i fati fæt’i horn waxra ndãni house tʃan uta ngu

I (1sg) n naja nuga geke nugi in na ha knee qunquri ɲãhmu to know mɑti jati pãdi to laugh wetskɑ laru thede leaf laphi ʎaqa ʃi leg/foot ikʃi tʃara kaju wa ʃinthe liver eel k’iwtʃa ja long sajt’u ma louse ɑɑtimi lap’a t’o cont.

96 Table A.8 cont. mouth kɑmɑ laka ne name tookɑ suti hu thuhu navel tsiipi kururu ts’ai neck kwetʃ kunka ʔjɨga new jɑnkwi matʃaqa ʔraʔjo night aruma ʃui nose jɑkɑ nasa ʃiɲu not ɑʃ xani hinã old thantha mãdã ndãʃxwa one kee maja ʔna rain wetsi xaʎu ʔje red tʃupika atʃhiwita thæni nthæni root nelwɑ saphi ʔjɨ rope mekɑ phala thebe to run tɬɑloɑɑ xala xithi nest’ihi t’ihi salt istɑ xaju u sand ʃɑɑl tʃ’aʎa ʔbomu to say sa mã to see ittɑ uɲxa nu shade/shadow ekɑwil tʃ’iwi ʃudi skin/hide kwetɬɑʃ xantʃi lip’itʃi ʃifri koto nts’ɯti small tsin xisk’a lænk’u lotʃi nlængu smoke pook xiq’i xiwq’i ʔbifi soil tɬɑɑl laq’a hai to stand sajt’a star siitɬɑl warawara tsʌ stone/rock te qala qarqa do maʔje medo to suck tʃiitʃii tʃ’amu tsɨt’i cont.

97 Table A.8 cont. sweet musq’a muʃsa u ugi ɲʔu

misk’i tail witʃ’inkha ts’ɨ

witʃ’inka to take wiikɑ apara hæts’i thick tilɑɑwɑ khusu lankhu lun- ndã pidi mpidi

qhu thigh mets nʃiɲã ʃinthe this ni aka nã nunã nuni to tie tʃinu ɲatʃ’a tuʔti ʔwæʔmi tongue laʃra xahne tooth tɬɑn ts’i water ɑɑ uma dehe what? tɬe kuna te who? khiti toʔo wide pɑtlɑɑwɑ phuʎtʃu nʃidi wind thaja ndãhi wing tʃhiqha hwa zahwa wood kwɑ lawa za yesterday jɑɑlwɑ wasuru mãnde you (2sg) t xuma nuʔi geʔe

98 Table A.9: Wayuu - Basque - Ngäbere

Family/ Language Maipurean/ Wayuu Basque/ Basque Chibchan/ Ngäbere

[ISO 639-3] [guc] [eus] [gym] ant xejuu s̥ inauri tʃinguri ŋaɲa arm/hand atɯna axapɯ beso bes̺ a esku kise kobɔ kude ash paliʔi auts̺ aus̺ tere erauts̺ ŋubrũ back asapɯ bis̥ kar tɾɤ big mijoʔu aundi kɾi bird wutʃii tʃoɾi nukuo nogwɒ to bite oxotta os̥ ka aginka bitter iʃ s̺ amin mikats̺ min-

gaits̥ black mɯtʃiija juulii belts̥ dɾune blood iʃa aʃa odol dori dorie to blow waawata awata aus̺ potu ais̥ eman suru bone xiipɯ eipɯse es̥ ur breast aɾuuwain atʃira bular ugats̥ bruku to burn (intr.) xota ere s̺ umin eraus̺ to carry alɯʔɯxa oʔunira eɾaman kareia wẽ

eʔika garaia child (kin term) atʃon ume aur s̺ ein ŋobriõ to come anta etori jin erdu duka xata nuke to crush/grind ajɯɯxa ankape oinas̥ pi irin tɯgu to cry/weep aʔjalaʔa muja mujẽ ŋetɾ to do/make aaʔinra aainxa aku- egin mike nomine

maxa dog erɯ s̥ akur tʃakur or nu nukrʌ to drink aapa eka edan ɲa ear atʃeʔe belari beari cont.

99 Table A.9 cont. to eat eka jan janaɾi mɾʌɾe kwet egg aʃuku araults̥ a araults̥ e mɾuke hudɒ

arauts̥ a eye oʔu begi ugwɒ to fall oxutta axutu eɾoɾi beko mʌbe nʌto far watt uruti aparte mendẽ neo mobe fire siki asema s̺ u s̺ ute s̺ uketa niʌ ɲugwɒ fish xime arain arants̥ a wa wau flesh/meat eʔiruku asalaa okela mede ŋotɔ ŋɾi fly xaʔjumulerɯ euli munkuo to give aapaa eman biẽ biãdre biẽ to go oʔuna joan nikiẽ noĩ ɾikia good an anami on s̥ ints̥ o s̥ us̥ en koĩ hair aʔwala walaʃi ile ilaje drũlẽ druɲen

dugwɒ hard itʃ gogor xume ɾibi

(s)he/it (3sg) ʃia nia aʃeʔeta aɾek ark uɾa iwe ko ni to hear aapaa ents̥ un adi belariɾa doĩ xaɾoa ɾu heavy xawat as̺ tun s̥ ama pisu dobʌko to hide ounxula es̥ kuta estali isilpe uke to hit/beat aʔjaata axata jo s̥ afratu doke mete toke horn oʔuwa adar house miitʃi epia etʃe gji xu

I (1sg) taja ni nik neu ti in aluʔu aaʔu ainkɯin n etan te de knee asapain belaun to know atɯxa jakin ga to laugh asiraxa leaf apana ori cont.

100 Table A.9 cont. leg/foot asaʔa ooʔui anka oin ŋotɔ ŋuɾe ulie liver apana gibel long mijoʔu lus̥ e ŋa louse mapɯi s̥ ori kũ mouth aanɯkɯ mutur kada name anɯlia is̥ en navel omotʃo s̥ ilbor s̥ il tuklɔ neck anulu lepo idun dʌro new xeketʃi tɯ beri night ai gau gaualdi arats̺ dew nose eʔitʃi s̺ udur mutur not nnoxo es̥ ɲã old laɯlaa adin s̥ aar kiɾa umbɾe one wane bat ti otʌiti rain xuja euɾi ɲu red iʃoo gori root ourala ereboil ero s̥ us̥ ter ŋotɾi rope xiiku apɯ lokari soka ɾʌbo to run awata kori nẽ salt itʃii gats̥ sand xasai ondar aɾea to say maa esan eran iɾaunts̺ i ɲe to see eʔra ikus̺ i gaĩ nikren toẽ shade/shadow ajoloxo eemiouʃi geɾis̥ a geɾis̥ pe its̥ al tɾo skin/hide ata paaʔata as̥ al laru kuata small motsoo xoʔuutʃon kos̺ kor braiko tʃi smoke amɯsain ke ŋubu ŋubuje soil mma lur dobo cont.

101 Table A.9 cont. to stand aʃaʔwala ʃaʔwata s̥ utieɾas̥ i tente nainkrʌ nunenko

nɯnaga star xolotsɯ is̥ ar argis̥ agi muke stone/rock ipa ari aits̥ eta los̺ a hɒ to suck atʃuʔla iɾents̺ i bose doin sweet pɯsiaa as̥ ukre es̥ ti gos̥ o mane tail asi bus̥ tan is̺ ats̺ to take aapawa artu ats̥ eman dẽ bu thick kaʔlee laɯktaa lodi thigh apɯʔɯ is̥ ter this tʃi tɯ na au auʃe ne to tie exitta lotu lokari soka mokodɾe tongue ajee mii mingain min tidɾo tooth ai orts̥ agin tu water wɯin ur niɔ ɲɔ what? kasa s̥ er s̥ ein dere who? xarai jaralɯ nor nork norts̥ u nire wide xerulaa s̥ abal wind xouktai ais̥ e murie wing atɯna ego egal wood wunuʔu ool s̥ ur egur yesterday soʔukalinka ali- ats̥ o xõdeni hɒdɾĩ

ikainka you (2sg) pia s̥ u s̥ uk s̥ uek mo ni

102 Table A.10: Highland Totonac - Khoekhoe - Galela

Family/ Language Totonacan/ High- Khoe-Kwadi/ West Papuan/

[ISO 639-3] land Totonac Khoekhoe [naq] Galela [gbi]

[tos] ant tʃaan ǂoLweLɾuHH guludu

ǂoLpeLɾuHH

ǂuLpuLɾuHH arm/hand maka maqa paq ǁʔõHãL !ʔoHHmH gia gija ash ɬka̰ ka̰ tsaHHoH gaapo gaʔapo back ǁʔãLLãH !huLriHH dudu big tlanka̰ kaHHiH lamo bird spuun aLniHH namo to bite ʃka nãHHãH goli gege bitter ʃṵṵn aLuHH mali black tsḭtsa̰ qa ŋǂuLLuL taro blood qa̰ ɬnḭ ǀʔaHoL au to blow !oHmL wu bone lukut ǂxhoLoHH kobo breast kṵʃmuu saHHmH mamata to burn (intr.) ɬku pasa smama kxaLLuH uku sora to carry lḭja̰ lii taHniL tide child (kin term) kam ǀoHHaH ŋopa to come mi haLLaL ǀxhiLiHH ado to crush/grind tʃḭnta tʃṵnta lat- xaHHiH waro diodo

tʃṵnta to cry/weep ta̰ sa tlawa aLaHH ari to do/make tʃitʃḭ tiLLiH kuHHɾuH aka dog kotua̰ aLLɾiH oHm- kaso

LaLɾiLL cont.

103 Table A.10 cont. to drink wa̰ qo aHaL udo ear takaa takampii ǂaLeHH ŋau to eat wa̰ ǂʔũLũHH odo egg qa̰ a̰ ɬua̰ a̰ t ! ʔuLwuHH ! boro

ʔuLpuHH eye mũLLũH lako to fall tapḭlḭ ɬtṵh ŋǂaLLaH tura far makat ŋ! uLLuL ŋ! uL- kaku

LuHH fire ɬkujaat ǀʔaLeHH uku fish skḭḭtḭ ǁaLuL nao flesh/meat liiwa̰ ǁaLLnH lake fly siimaawa kṵʃta̰ ǀĩLĩLnaLL ǀĩLĩL- guʔupu

naHH to give maasta̰ a̰ ʃtaq maLLaL hike to go a̰ !ũHũL tagi good tlaan !ãHĩL loha hair ja̰ h tʃḭʃḭt ǀʔũLũL hutu gogo hard paɬa̰ !aHHɾiH togoi

(s)he/it (3sg) wa̰ ʃla̰ ǁĩLĩL iLL una muna to hear qaʃmat a̰ qawaana ŋǁãLLũH ise heavy tsinka !oHHmH tubuso to hide tsḛḛq sãHũL masi waʔihu to hit/beat ka̰ ʃi slama nik ŋǂHaHHu hoHHoH ŋapo tero dubu horn a̰ qaloqot ŋǁãLLãL tadu house tʃikḭ oHmL tahu

I (1sg) taL ŋohi in ŋ!ãLãL maraba knee ǁoHHaH pupuku cont.

104 Table A.10 cont. to know ka̰ tsii ka̰ tsiinḭ ǂʔaHHnH nako to laugh liitsḭḭ ãLĩL dohe leaf tṵwa̰ a̰ n ǂaLeHH ŋǂaLɾeHH soka leg/foot tantuun tuwan ŋǀuLLuH ǂʔaHiL dohu liver ɬua̰ ka̰ ka̰ ãHHĩH ãLĩH dane long ɬma̰ a̰ n ɬma̰ a̰ n hõLLãL kaku louse skaata uHɾiL gani mouth kiɬ kaɬ aHHmH uru name tapaakuwiit ǀʔoLnL roŋa

tukuwinḭ navel tampṵs suLniHH woti neck piʃ !ʔaHHoH tolo new ʃasaastḭ ǀʔaHsaLL momoane night tsḭḭsnḭ tsuxu putu nose kinka ǂuLLiH ŋunu not niinta̰ taLmaL wa old ʃapuʃkṵ ʃatlanka̰ ǀoHHɾoH kaHHi- pereki teka

lakuan HɾaL one a̰ qtum ǀuHHiH moi rain siin ǀʔ‘awiHH ǀʔ‘apiHH muʔura red tsutsoqo ǀʔaLwaHH ǀʔaL- sawala

paHH root tanqaʃeeqa ŋ!oLmaHH ŋutu rope tsuHHuHɾiLL gumi to run qosna !kxoHHeH magogora dodagi salt matsat ǂʔoLoL gasi sand kukuj ǀoLwaHH ǀoLpaHH dowoŋ to say wa mĩHHĩH temo tʃarita to see ṵkʃiɬ mũLLũL nano cont.

105 Table A.10 cont. shade/shadow sqa̰ a̰ qna̰ soHmL dadaru riho skin/hide ʃuuwa̰ kxoLoHH kahi small a̰ ktsuu ʃasqa̰ ta̰ ǀxhaLɾiHH tʃeke

kaLLaLɾaLL smoke hinḭ ǀaHnL thopo dopo soil tḭja̰ t !huLLuH tona to stand taaja tatʃoko mãHHãH maʔoko star ǀaLmiHHɾoHH ŋoma stone/rock tʃiwiʃ ǁhoLLmH ǀʔuHiL teto

ǀxhoLmHH to suck tsṵtsṵ ʃookua̰ tii ǀoLmHH sodu sweet sa̰ qsḭ ǂxhoLnHH loha muti tail stahan ǂʔaHHɾeH pego to take tʃḭpa uHHuH dehe thick tʃa̰ a̰ tlanka̰ ŋ!aLuHH ǁʔaLuL dodiao thigh tantiin ugu pugili this neLLeL manena mana to tie !aHHeH piliku tongue sḭḭma̰ qa̰ a̰ t naLmHH lade tooth tatsan ǁũHHũH iŋi water tʃṵtʃut ǁaLmHH ake what? tuu tukṵ taHHeH mãLãHH kija kia

taHHɾeH who? tii tikṵ taLiHH haHHmH nagoona

taLɾiHH wide tiitlanka̰ haLLɾaL hõLLãL ŋoha wind ṵṵn ǂʔoHaL paro wing paqan ǁaLLwoL ǁaLLpoL wood kḭwḭ ŋ!aHHoH gota cont.

106 Table A.10 cont. yesterday qootana̰ ǁʔaLɾiL kagunugo you (2sg) wḭʃ s ts ŋona

107 Table A.11: Mapudungun - Western Highland Purepecha - Woods Cree

Family/ Language Mapudungu/ Tarascan/ West- Algic/ Woods Cree

[ISO 639-3] Mapudungun [arn] ern Highland [cwd]

Purepecha [pua] ant koʎaʎa ʃɨɾuki iiðik arm/hand lipan lipaŋ xakhi spitwan spiton tsi-

htsii ash tɾufken pihkotiiw back fuɽi kopaɽakwa spiskwan big fɨtʃa kheni misikitii kihtsi bird ɨɲɨm kwini piðiisiis to bite ɨna katsaɽi maakom maakwaht

tahkwaht bitter mukɨr khameɾi wiisakwan black kurɨ tuɽipiti kaskitiiwi blood moʎfɨn juɽiɾi mihkw mihko to blow pimu phunita isiðowii bone foɽo uni skan breast mojo itsukwa toohtoosim to burn (intr.) ɨj ɨje ɨj aphaɽi kuɽi tiʃara pasis pasitii pasiso to carry je pa tahkon child (kin term) pʉɲeɲ watsɨ wapha awaasis to come aku kʉpa xuɽa takosin to crush/grind trana tʃhata uɽu saamakatah

pinipotsikii to cry/weep ŋɨma wajlɨ weɾa maato oht-

sikawaapi to do/make dewma uni toot htaa dog ʈɽewa witʃu astimw tiim cont.

108 Table A.11 cont. to drink pɨtu itsɨma minihkwii ear pilun kutshɨkwa htaawakaj to eat in thiɾe miitsi po miitsiso egg kuɽam kwaʃanta waawi eye ŋe ɲe skiisikw tsaap to fall naw ʈɽan poksata kawipii pahkisin far kamapu jawani waahðaw fire kɨʈɽal tʃphiɾi iskotiiw fish tʃaʎwa kuɽutʃa kinosiiw flesh/meat ilo kwhiɾipeta wijaas fly pɨlɨ tinti ootsiiw to give elu intsku miið to go amu niɾa itohtii good kɨme ampaketi miðo hair loŋko kal ʃawiɾi iskw stakaj hard jafɨ tʃopheɾi maskawaa

maskawisi

(s)he/it (3sg) fej ima wiiða to hear aʎk kuɽa piiht piihtaw heavy fane khwetsape kwasikowi to hide eʎka elka xiska xirikuɽi kaaso aakoomo

kaat to hit/beat mɨʈɽoŋ ata thawa taaw pakamahw

pasikiðaat horn mɨta ʃɨwankwa iiskan house ɽuka kt’a kumantʃikwa waaskaahikan

I (1sg) intʃe xi niiða in ponwi punwi o ihk knee luku htsikwan cont.

109 Table A.11 cont. to know kim miti kiskiiðiht to laugh aje aj teɾekuɽi paahpi leaf tapɨl tʃkhuɾi pak niipij leg/foot tʃaŋ namun tsikathakwa kaat skaat sit

xantsɨri liver fukuɲ ke tawaʃɨ skon long jothi joɽati kinwaa iskwaa louse pɨʈɽaɽ tɨn ampuʃɨ ihkwa mouth wɨn kwoniiw ton toon name ɨj wiihð navel tupu tisij neck pel kwajaw new ne ximpa osk night pun ʈɽafujan tʃuɾikwa tipiskaa nose ju uɽi kot not la no iikaa ma maa old fɨtʃa kutsɨmiti thaɽepiti one kiɲe ma piijak rain mawɨn xanikwa kimiwan red kolɨ tʃaɽapiti mihkwaa root folil ʃɨɾankwa watapij otsipikwan rope zef ʃɨntaɾi piiminaahkwaan

piisaakanaapij to run lef nekul wiɾia pahtaa salt tʃadi ithukwa siiwihtaakan sand kujɨm kutsaɾi ðiikaw to say pin aɽi itwii to see leli pe iʃe aapi waapi shade/shadow ʎawfeɲ ajwiɲ kumanta tsikaastiisin cont.

110 Table A.11 cont. skin/hide ʈɽawa sɨkwiɾi pahkiikinw sakaj

asakaj small pitʃi sapitʃu apits smoke fiʈʂuɲ ʃɨɾata aapahtii soil mapu etʃeɾintu ask to stand witra xawaɾa ananti kaapawi star waŋɨlen xosku atsaahkw stone/rock kuɽa tsakapu asinij to suck fotʃiz ʃuɽundi ðowam ðowaaht sweet kotʃi teɾi siiwaa tail kɨlen tʃheti sowaaðwij

sowaaðiwij to take tu xupika otihtin thick xuntu kispakaa thigh ʎike tsikaxtakwa pwaam this fa fej i awa ooho ooma to tie pɨɽon xo kipwahpit tahkopit tongue kewɨn khathampa tiiðanij tooth foɽo ʃɨni iipit water ko itshɨ kam what? tʃem ampe kikwdy who? iɲej ne awiina wide ko mahki wind kɨɽɨf taɽjata ðotin wing mɨpu kaʃuɽa tahtahkwan tahk-

wan wood mamʉʎ tʃkhaɾi aahtikw aaskw iht yesterday wija otaakosii you (2sg) ejmi thu kiiða

111 112 Table A.12: Navajo - Highland Popoluca - Ambulas

Family/ Language Eyak-Athabaskan/ Mixe-Zoquean/ Sepik/ Ambulas

[ISO 639-3] Navajo [nav] Highland Popoluca [abt]

[poi] ant hahtsuk arm/hand kaanL laʔL haʔŋku kɨʔ taamba ash ɬeeʃLtʃ’ihL bawu back tuʔɲi mbutŋgul big tsoL thiiHl tsaaHz mɨh apakələ nəmaan

tirapa bird tsiHdiiL hɔn api to bite haLʃ was saŋgul ti bitter tam ɲəŋgi black ʃĩĩHɬ jɨk ŋgələ blood txiLɬ wəɲ to blow joLɬ sus pup japəv dʑur bone ts’iLn pak apa breast ɣitL nuunu muɲaa to burn (intr.) k’ããHɬ hips jaan to carry ʔaaHɬ səlakət child (kin term) ʔaHɬtʃhiHniL maanɨk ɲaaŋusə ɲaan to come kaaHɬ miɲ jaa to crush/grind tʃĩLʃ k’aaHɬ haʔp ŋgətɲa to cry/weep tʃaLh weh ŋgəra to do/make niiHl liiHɬ siLh wat ja kuttakn dog ɬeeHtʃhããLʔiH tsimpa waasa to drink tlhĩĩHɬ ʔuk k ear tʃaaLʔ taatsɨk waan to eat ɣĩĩHɬ kuʔt wiʔk k cont.

113 Table A.12 cont. egg kaʔnpu ŋgək eye naaHʔ məni to fall taLs tsut akərə far huʔmɨ səknaa fire kõLʔ huktɨ jaa fish ɬooʔH tɨʔpɨ flesh/meat tsĩʔL maaji kwaami fly hɨɨhɨ saat to give ʔaaHɬ tsiʔ to go kaaHɬ nɨkk j good wɨ jəknwun mitək hair tsiiLʔ waj nəmbə hard tɬh’iLs kamam tirapa atku

(s)he/it (3sg) biH hɛʔ lə də to hear ts’ĩĩHɬ matoŋ vəknwu heavy taaLz taaHs tsaʔksɲi ak to hide t’ĩĩHɬ ʔĩĩHɬ jam neʔk paaku to hit/beat txaLɬ sããHɬ tʃhiiHl koʔts vijaa jaanəpək

ŋgətɲa horn deeLʔ kaara kaara house kiLn tɨk ŋga

I (1sg) ʃiH ʔɨtʃ wunə in piiLʔ mɨ awula knee koLd kɔɔsɔ kwati to know tʃhĩĩLh zĩĩLɬ sĩĩLɬ hɔɔdɔŋ kutndəŋ to laugh tɬohL ʃik waaŋgi leaf t’ããʔL ʔaj miŋga mpija

ŋgaŋga leg/foot tʃaadH puj maan cont.

114 Table A.12 cont. liver zitL tsɔɔgɔj wutmi long neezL ts’oozH jagats səməɲ louse jaaʔL ʔaʔwat ɲəmu mouth zeeʔH hɨp kundi name ʒiʔL nɨɨji jə navel ts’eeʔH tuunu ŋgwaraandʒən neck k’osL ʔɨskɨ kwaalə new niitL hɔɔmi kulə night thɬ’eeʔH tsuʔ ŋgaan nose tʃĩihH kiiɲi taama not ɟa kaapuk old hahL txihL tsaami ɲaavə ŋgwalepa

ndʒaaba one ɬahL tuum tjuum kipa nak nakurak rain txĩĩɬH tuh maas red tʃhihL tsabats ŋgwaavə root tiitsɨk məŋgi rope tɬ’ooɬH baagwi

balədʒawurək

sasən to run ɣoɬL t’aʃL tɬ’eeɬL pɔj pətə salt ʃĩĩhL kaana sol kus kusɲət sand seHiH pɔʔɔj jawusa to say niiɬL nɨm na w to see ʔĩĩɬL tseeɬL ʔiʃ v shade/shadow kumuŋkɨ apaŋ kajəkni skin/hide naaka səp small jaHʒiH ʃuutju makwal ɲatkwal smoke ɬitL hɔɔkɔ jaatɲə cont.

115 Table A.12 cont. soil ɬeeʒL nas kəpmaa to stand zĩĩɬL teen t star sõʔL maatsa kun stone/rock tsheH tsaʔ matu to suck t’oɬL tshoɬL ts’õsL tuʔt ndʒəlip sweet paʔak lisək ləsək tail tsheeʔL tuʔts gəni to take ʔaaɬH ʔaɬL pãsL mats kəra ɲəgəl thick pɨm thigh daawa this tiiH tiiHtiH jɨʔp ani kən to tie tɬ’ooɬH tsen ŋgi tongue tsooʔL tɔts təknajələŋ tooth ɣooʔL tsɨts nəmbi water thoH nɨʔ kulak ŋgu what? jaahH tjii tji samu jaŋga who? ʔii ʔi kijadə wide wɨti wind saawa wimut wing t’aʔL sah pajək wood tʃhiʒL kuj mi yesterday matɨʔk nalə you (2sg) niL mitʃ ɲən mən

116 Table A.13: Mískito - Shuar - Northern Emberá

Family/ Language Misumalpan/ Jivaroan/ Shuar Chocoan/ Northern

[ISO 639-3] Mískito [miq] [jiv] Emberá[emp] ant papu tʃinkiamar kaabudɨd’ɨ arm/hand mita ewex kuntu hʉwaa ash jampus junkunim ntarra back nina tuntup ẽŋkarra big tara jaxpa maixa uunt timian oaibʉa bird tnawira piawira nanamtin tʃinki ĩmbana to bite sam esar kai bitter taxpla japa oregea adsea black siksa mukusa phaima blood tala numpa oa to blow pub hãrãmpuai bone dusa ukuntʃ’ bɨbɨr’ɨ breast tialka lama muntsu netsep hu dhodaa

lamkuka to burn (intr.) klau kapãɾ ekemak eradrɨbɨr’ɨ to carry briw kwaxk xuki ethei eneĩ child (kin term) lupa lupja kau oarra to come bal uru to crush/grind wau akb kap bʉai bhai to cry/weep in ut hĩaĩ to do/make dauk oi dog jul jawa aiʃmank usa to drink di umar dhooi ear kiama kuiʃ kʉwʉrʉ to eat pi nutʃer neekoi egg mabra nuxint teeri mmʉ cont.

117 Table A.13 cont. eye nakra xii dau dhau to fall kaux iɲaaɾ bhaai far laiura lai wixka jaxa timiai arant tmʉ̃ fire pauta xii fish namak bheda flesh/meat wiina namank nedʒaraa fly kukas kaap ĩŋganana to give jab pruk diai dhiai to go w wap we õaĩ good jamni aitani penker papu bhia hair taawa intiaʃ ure buda hard karna katsuram kakaram zareabʉ

tee

(s)he/it (3sg) ba witin au nii idʒi ãndʒi to hear wal ʉ̃r’ĩ heavy wira kixin sɨggʉa to hide jukuk maprik uuk mẽrai

yukuw to hit/beat pruk karai boropei apor-

reaja horn masas aintiaku katʃo house utla watla katʃu dhe

I (1sg) jaŋ xea mʉ̃ in ra bilara nam num idʉ knee lula tikiʃ chir’ãmboro to know kraik neka kaoai to laugh kik iphidhai leaf waxja waxa nuka kedua leg/foot mina jaman wajata maku nawe bhaakara hĩr’u hĩru cont.

118 Table A.13 cont. liver auja akap thaãrɨ long jarika wixka esaram drassoa louse khua tema t’mbarabʉ̃ mouth bila wene itte name nina naa tr’̃ navel klua untutʃ kmʉã̃ neck nana kuntux otʃtʃiru otʃiru new raja dʒiwidi night timia tama kaʃi dimisi nose kakma nuxi kẽmbʉ not apia ras apu ẽ old nanaswan untatʃ uunt aseabi dr’õã one kum kumi bhar’ã enena rain li kue red pauni kapaku dʒipurru root plamaja wakia tserup nekarra rope awa jarank tʃapik’ hũŋkara to run plap tseken pirai salt saal katʃi sau wee th’ãã sand nɛikim ĩmbʉ to say wi titin harai to see kaik unui akʉi shade/shadow dista distika kirit mikintiar nẽmbai kũrasare skin/hide taja waimus nuap sexik ẽe ee ezake small sirpi lupia utʃi utʃitʃ’ iʃtiupatʃ kaabeeabʉ

maar’ĩmbʉ smoke kiasma mukuint’ kouwa soil tasba nunka udaa to stand bu wahas nũja cont.

119 Table A.13 cont. star slilma lusero stone/rock walpa walpaja ak- kaja katsuram mõŋgara

baika to suck utb utk thãnsoi sweet damni k’ũa tail wuaika uxuke dru to take alk edai thick twakni kampuram nu- thãmmabʉ

param thigh kuxma maku bhaakkara this na naha xu zaabʉ to tie wilk xinkia ligai h’ndhã hɨni tongue twisa iniai kẽrame tooth napa nai kidha water li laja entsa baidho what? ba warim k’ãre who? ja jamna jaa kaidha kai wide jahpa wankaram heddea wind tnawa majai nase n’ãũra wing tnawa nanap i wood dus taat bakuru yesterday naxwala jau nuwena you (2sg) man ame bʉ

120 Table A.14: Bukiyip - Northwestern !Kung - Greenlandic

Family/ Language Torricelli/ Bukiyip Kx’a/ Northwest- Eskimo-Aleut/

[ISO 639-3] [ape] ern !Kung [vaj] Greenlandic [kal] ant ‼uH‼uH unijaaluk arm/hand ‼hãLL gǁaLoL talɜq assak ash gǁoLLhaL arsat back !oMbeM ǁxaMũH qatsiɣak tunu big !aMŋH aŋisɔɔq bird tʃʔaHm tsimmijaq to bite !ʔaMeL bitter g!aMʔaL suŋɑnnip black dʒoM g‼xʔaMiH qɜnnɜttoq blood n‼oMʕluH aak to blow maʕLL anɔɬɬɜ bone !ʔuH saanɜq breast ʃaHm kuMuH amaamaɣaq ivi-

jaŋɜq to burn (intr.) kuMʔuH ‼ ʔhuLL- ikuwala ikuma

biLL to carry tãLniL assɑttɔ ɜssup child (kin term) daLbaL mɜɜʁɑq qitɔnnɑq to come g| eL ǁxaMiM pi

ǁxaLiL to crush/grind ipissa mɔʁɜɜ to cry/weep dʒʔŋH qija quɬɬili upip to do/make oLL kuHluH ili pi sana

gǁaLLhoL dog g‼hoMeM qimmɜq to drink ʃŋLL imiŋŋɜʁ cont.

121 Table A.14 cont. ear |ʔhuHiH sijut to eat mH tamɔ tamulup egg n!uLuM mannik eye g|aLʔaM isi to fall g‼aL ɔɬɬu far !ʔaL uŋasip fire daLʔaL innɜq fish g! oLloM ! hãLL- aalisaɣɑq

nuL ǁʔaHuH flesh/meat ‼ʔhaLuH ǁhaM uwinik nɜqi fly dʒoLãMdʒoLãM nivijuɣak

ndʒoLãLndʒoLãL to give neʔLeL |aLʔaM tuni to go uH xuH pisup good kaL hĩM hair !xʔuHiH nujɑq mɜqqɔq hard ǁxʔaMoM niŋŋusɔɔq mat-

tusɔɔq

(s)he/it (3sg) haLL kaM una to hear tʃaL tusɑ heavy tiLLhiL to hide kaʔLLmL tɔqqɔ to hit/beat n!hoM n‼aMʔmL unatɑ pɜssup horn !huH nassuusɑq nassuk house n!aMoM iɬɬu

I (1sg) maM uwaŋa in n!HŋH mi knee g!xoLLaLL sɜɜqɔq to know ʔn‼eLLhiM cont.

122 Table A.14 cont. to laugh ʃiL |ʔhiH iɬɬɑ leaf g! oʕLLaL pilu mulɜqut

geLǁxʔaHeH leg/foot !oL ǁmH ǁʔhmL isiɣak niju saanɜq liver ʃŋH tsiŋuk long g‼aMʔŋM sivisɔɔq sivisu tak-

isɔɔq louse dʒʔŋH kumak mouth tʃʔiH qanɜq name !uH ɛtɜq navel n!oLʕmH qalasɜq neck ǁaLLŋ tɔqqusɑɑq quŋasɜq new dʒeLLheL nutɑɑq night g|uH unnuwaq nose tʃxʔaHŋ qiŋɑq not |oMaM ŋŋilɑq ŋŋitsɔq old g‼aLLhaLL pisɔqɑq utɔqqaɬɬak one n|eLʔeL ataasɜq rain gǁaL sijaluk red g!aLeM root nǁaLiL ǁaMliH sɔɬɬak nijulutɑq

maŋŋɔq rope n‼oLLʕ tʃoLʔoL aɬɬunaasɑq bɑddu-

uni nusukkɑʁut to run ‼ʔhuLaLLm qɔɬɬuula ɑppap

paŋalip salt guHiH taʁatsut sand kxʔaL sijɔqqat cont.

123 Table A.14 cont. to say koLeL kuMjaM

kweLeH to see hŋH qanimɔ taku shade/shadow ǁʔhoLãL ! haMi- alaŋŋɔq taχχɑq

MãM skin/hide n|oM ‼oLaM amɜq small tʃʔeMmaL quttsɔq smoke ʃuHleL nǁoLʔoL ‼ pujɔq

oLLhaL soil kxʔaL to stand gǁaL nikɔffa star ‼ũLL uɬɬɔʁijɑq stone/rock n! mL | ũniH | ujɑʁak

xuLmM to suck kaMloM ʔnaʕLL- miɬɬuwɑ

luH ‼ʔhũL sweet n|ŋLʔŋH inɜqunɑ tail ǁxoMeM pamijɔq to take nǁŋL avalaap thick gǁxʔaLL issusɔɔq thigh uppat this ŋMŋL manna to tie ǁʔŋL qilɜ tongue d’hLaLliL ɔqɑq tooth tʃʔaMoM kiɣut kiɣutaasɑq water gǁuH imɜq what? mMtʃaH suna who? mMdʒeM kina

mMdʒoMeM cont.

124 Table A.14 cont. wide khaLLlaLL nɜʁutusɔɔq

nɜʁutɔɔq wind ǁoʔLLliL anɔʁi wing g‼haLbaL isɑʁɔq suluk wood !oMbeM qisuk yesterday ǁhaMeM ippassɑq you (2sg) baL iɬɬit

125 Table A.15: Burushaski - Sentani - Terei

Family/ Language Burushaski/ Bu- East Bird’s Head- South

[ISO 639-3] rushaski [bsk] Sentani/ Sentani Bougainville/

[set] Terei [buo] ant pilili numbɛ kunna arm/hand ɕak mə aku ash phet odowaj ondowaj peu back waldes mokodu nomo big nju kabam eke bird ʈʂen aje peɾekupa to bite gaʈ akə kaku bitter qaqam fearə apaɾa black matum nokomom muu muuŋ kumi blood multan ha oki inna to blow dVɕaragin mawəsi uupu bone ten bo kaana breast mamu nimə nutu to burn (intr.) ɣol hon ugu muu to carry balda a mua child (kin term) dulas fa ite ruŋ to come tsrum mə u to crush/grind dVgurgun sark jaleizale koŋ to cry/weep her dimə piapia to do/make t ko moko apa ee dog huk joku maikuna maimai to drink men anəsi ɾoi ɾai ear ltumal tumal ankɛj nume to eat khana anə tamu egg ʈiŋan ro uoɾo cont.

126 Table A.15 cont. eye ltɕi idoko noine to fall bal duw duwə u far mathan ahaw itsigou fire phu i oguai fish tɕumu ka iiana flesh/meat tɕap məlɛm mootina kaaku fly phen bɛnso kuɾuɾu to give u je i ij o to go tsrem ə pe good ɕua foj meɾagu hair ɣoyaŋ uma ma ii hard ɖaŋ tɕiɖiŋ kuraʈ poɾogupa

(s)he/it (3sg) ne mo se na ako eko to hear dVjal bodo ɾuge heavy tsulum əhə mokinatsi mokina to hide urk jane mina nina to hit/beat del habə wə to horn tur house ha imɛ opo

I (1sg) dʑa da ne in ule na nə knee numus nuŋus osələŋ mii utoutoia to know hen moɾu maugu ono to laugh ɣas hadə kukuɾe leaf tap fɛ oto ota leg/foot huʈis odo kaga liver ken phuhɛ nunno long ɣosanum həki itsi kiɾikiɾitu louse kharu mi otsi cont.

127 Table A.15 cont. mouth xat əwwa muŋ name jek kɛnə miŋ navel su bulɛnde ituge neck ʂ kambi hodo ou new thoʂ hasil nəmə ɾiɾo night thap denijaj muuɾo nose muɕ joj keŋ not a ai aj bam am tou old akabir dʑaʈ majen bɛna dele nale uni one hek hen əmbaj nonumoi rain dau ja muɾu red bardum hajsaj iipi root tsereʂ kambu kage rope ʈete haa kuuku piaɾuɾu to run hue gaarts haka hebɛlə kuɾo salt baju hɛ tsiitake sand sau kiho uutsia to say sen ələ aa o ɾaɾe to see jets ədə ŋ shade/shadow nal nulu uɾa skin/hide baʈ wa katu kagu koo small tɕojon tɕuɣun kəndim kua otsinai ɾikiɾiki smoke thas jokonim iito soil tik burdi kani mootina kaaku to stand ɣe həbə ita star asumun dalo kaipa stone/rock dan duka aru antsiɾu makat-

siɾu to suck moɕqo jophoiphoi mutu cont.

128 Table A.15 cont. sweet njam ju muɾiɾatsi muɾiɾa tail ɕilan dəmə mogu to take jan dow dəw jado ɾo aapu koi thick daɣanum dundum muutu thigh ʂak bɛdə moku this khene khin dakə bɛkə to tie phus riineko riindim ogopo uku tongue juŋus few meɾe tooth me idəha kompe water tsel bu tuu what? botan bo be maka ou oi paake who? men amen hinə pa wide ɕoqum kɛw kəw kamagatsi wind tiʂ alu kakagu wing galgi nodo kupo wood hun o kui yesterday saaʈik wena togega temu you (2sg) un wa ɾo

129 Table A.16: Macushi - Ap Ma - Páez

Family/ Language Cariban/ Macushi Ramu-Lower Paezan/ Páez [pbb]

[ISO 639-3] [mbc] Sepik/ Ap Ma

[kbx] ant miʔkɨ jili kxãʔkxã kxãʔng arm/hand jemeʔkon miaʔ lapoŋ kuse ash ʁunɨpɨ woka kxuuts back mopo lomot tsinz big kuʁeʔnan okkai ɲaam wala maxkj bird toʁon awon βitsyakwe βixtsya to bite jeʔka ma waʔkj bitter maiʔ epo jax black ʁikutun bund kxũtsy blood mɨn sɨlɨjen ee to blow ipuma kupol putx bone jeʔpɨton amo djiʔtx breast manʔtɨ tomboŋ petsyu to burn (intr.) eʔpotɨ pɨndat kanb to carry pintɨ tupa at jaat jakx child (kin term) manon mumu be ntsyiʔk niisa to come iipɨ sila jux to crush/grind waʔtsxu ũʔkx to cry/weep liil ũʔne to do/make koneka motɨl βit jũu dog aʁimaʁaaka ludɨma alku to drink jenɨʁɨ ma tundj ear panaʔto kombe txũʔwe to eat entomoʔka ma ũʔ egg pɨmoi mbui zits cont.

130 Table A.16 cont. eye enu laam jaɸj to fall enaʔ mapɨl wete saapaxkj far aminke baaŋ xjuʔx fire apoʔ pɨndɨma ipj fish moʁoʔ lamaŋ wendj flesh/meat pun saʁaʁu laam tsyitsy fly eʁewe lemund ɸĩsy ɸjnex to give itɨʁɨ lana ũsu to go wɨtɨ ela uʔx good moʁɨ jawulo ew hair siʔpo wale djkxas hard saʔman bɨlɨŋ wexj

(s)he/it (3sg) mɨɨkɨʁɨ mɨba man mɨn kjãa tjãa to hear wẽsẽʔx heavy amɨiʔne dɨnaat dux to hide enomɨ ka sũtx to hit/beat jal pekwe uji uxka horn ʁettɨ kom katsyu house wɨttɨ padɨma jat

I (1sg) uuʁɨ nɨba andj ũʔkwe in po min lin gin te ka ga knee jeseʔmu amboŋ xjũkwet to know eputɨ los xii xiju to laugh siʔsiʔ laal syixka leaf jaʁe sɨpan ets leg/foot ʔpu si sɨpaŋ tsyinda liver teʁeʁɨʔpɨ ewoŋ meʔkj long kusan somo xjuʔx louse ajan bunduma ẽs cont.

131 Table A.16 cont. mouth mɨta tin juwe name jeseʔ junduma jase jese navel poni betɨgum syanb neck pɨmɨ kotom tjxikj new amenan kakaan uʔse night ewaʁon bɨdan kus nose jeuna puŋ ĩts not pʁa pepɨn miɲa mee old ipoipɨ poʒ wao mom ĩisy txẽʔxsa one tiwin boa teetsy rain konoʔ lɨndɨma nus red suju dɨŋ bex root ikaʁa ladɨma watse wetse rope jewa alo wes to run ekaʔtumɨ mondutɨl wuwuu salt pan sim nenga sand kɨsaʔpan munʒ muse to say taa mamo kjãʔxĩ tjãʔxĩ to see eʁaʔma ux txeng shade/shadow bunduk psyũ skin/hide piʔpɨ potom kxaʔtj small simɨʁikɨ buŋo leʔtsy smoke poai ax soil bɨn kiwe to stand teaŋa juxu tisyi star kaiwonoʔ siʁɨkɨ lu ã stone/rock tɨ lambat kwet to suck juʔna jaxpe tsyandzya sweet aʔkiʔku mɨnaat nyusya cont.

132 Table A.16 cont. tail jaukɨ gapo menz to take moʔka lɨtɨna uwe thick ɲaam tsyal lepxj thigh ipeʔ xiʔnbe this seeni mɨseʁɨ gajen majen dajen naa to tie ewaʔtɨ tund tongue nu ɲo txune tooth je wa kiʔtx water tuna mɨdɨma juʔ what? ɨʔ ʒuma kĩx who? anɨ looma kim wide aweneʔ ɲaam tape wind asitun seʔman balamb wexja wing japɨʁi pa wood jei mop ɸjtũ yesterday koʔmanpʁa budan ũʔna you (2sg) amɨʁɨ wuŋa indj iʔkwe ixtsya

133 Table A.17: Wichí Lhamtés Vejoz - Sandawe - Xibe

Family/ Language Matacoan/ Wichí Sandawe/ Sandawe Tungusic/ Xibe

[ISO 639-3] Lhamtés Vejoz [sad] [sjo]

[wlv] ant siwan̥ as pãHbuL arm/hand tkwe tl’ũũ ɢɑɹ ash !’uHphaL filiŋ back tow̥ u dʒandʒa fidʐa big wefw baLʔ meHeH ɑm ɑmbu bird ʔafwentʃe thiwLiH ɢ̤ asχ to bite jukwaχ ǀ’ĩLkeL ɕiɑm bitter tapaj tl’weHeH ɢosqʉ̃ χɑq black tʃalaχ k’aŋk’ara jetʃĩ blood wʔijis !’aLaLkaL ɕiŋ

ǁ’eHk’aL to blow jafwit ǀnuLphaH ǀnũL- fitʃɑ

phaH bone ɬetʃʔe ɬile !iH girɑ̃ breast tʔate ǀ’iHnaL xuɣun to burn (intr.) lakʔe ǂnats’ indɘm to carry tʃoχ tiLkweH ʉzɑ child (kin term) ɬos ɬose ǁno ǁnoo dʒi to come nom ǀiH ǁweLʔoL dʒi to crush/grind itʃufwaχ noHweH ǁ’ĩLkheL midʑilə to cry/weep afwɬi keL k’eHeH qors qorso to do/make ʔiwuje ǀniLiLraL ǀnweHeH ɑr ɑrə

ǀnuLǀnuHuL dog asinoχ kaaka njoʁũ njunʁʉ̃ jindɑʁũ cont.

134 Table A.17 cont. to drink jo jojex ts’eH emi ear tʃʔuteʔ keke sɑ̃ to eat ʔek tefw mantʃha dʒə egg ɬetʃʔe dḭʔa umʁɑn eye teɬu ǀweHeH iɑs̤ to fall toɬpho !’aHweL tʉh tʉɣ tʉɣə far atufwej ʔiHthaL ɢor fire itoχ ǁ’ĩ tuɑ fish wʔahat somba njimʁɑ flesh/meat tʔisan ǀnĩL ieɹl fly ʔataq ǀãLtiLmaL dʉrvo to give po bʉ to go jik ʔãHãL jɑf jɑv jɑve good is ɬaH ɬaLuL ʃiẽ hair towʔulej ts’e !’ũ finh hard tʔen ǂnaŋk’o do

(s)he/it (3sg) ɬam hesu hewe ər to hear lot kheHʔeL doəndʑim

dœndʑim heavy nitʃujhat ǀeeka ʉdʒən ʉʒĩ to hide skat ǁwaL to hit/beat ʔjaχ lon khaLaL lalata fier fierə tand

mak’ase horn totʃʔe tlaHnaH uixə house laʔwet khoLoL bo

I (1sg) nʔɬam tʃi bi in hi kiLtaL t knee topʔuɬaq kẽH bux to know jihanex miLniLk sɑ cont.

135 Table A.17 cont. to laugh tʔistʃej k’oLjoH !neLeL indʒi leaf lawule ǀaLaL avχ leg/foot ɬetʃe patʃʔu ǁhaLtaH bətk liver totonek thaHsiLnoL fahun long pitaχ magandʒa ɢoəɹlmin louse ɬaʔ maLǀ’aL tɕhiɣə mouth tokʔaχ ǂnum ɑɴ name ɬej ǁwaL gəf navel totsʔaq ts’ũLbuL uluŋ neck wʔu k’wẽL tʃorʁo new ǁa itʃə night hunatsi !ĩHts’iL djovr jemdʒ nose lan̥ es ǀnaHthiL ovur not ha hi tsheL ts’e ɑqw vɑq vɑqɑ old lawemek thalokw ǀ’aLǀ’ãL fə sait sɛt one unu ts’eHxeL əm əmkə̃ rain iwumtʃoʔ tl’wãHãL ɑʁɑ red itʃot !’iLʔeH fəɹgiẽ root lafwetsi !nĩLĩ dɑ rope nijokw kuL fəta to run thaL fəkʃ fəkʃi salt nisuj ʔuɬi davsun sand alena xoLxoH bioʁũ njoɴũ to say juq boL sə zə to see ʔwen ǀʔee sɑv tɑ shade/shadow hipeɬ biLrĩHbiLriH ʔeʔõ həmlin

ǀ’õL skin/hide totʔoχ kelemba soqw small ɬemsas ts’oL ɑdʒi cont.

136 Table A.17 cont. smoke tetsaχ ts’uHk’aL ɕiaŋŋ soil hun̥ at ǀ’iHnaH na to stand taqasit ɬeHeH ji je star qates ǀnoHwãH ʉʃiχɑ uʂχɑ stone/rock tente diin din vɘɣə to suck hitsʔe ǀ’uLkuL gutʃi sweet nitsot maL ɑmtiŋ tail totʃos ts’waLaH unɕykhon to take tʃem khuL siH tl’aLaH ɢɑ thick tatʃʔupwaj boʔe dʒirɑm tømĩ thigh toɬetʃe dziLphaH this toχha hee ər to tie inajhit hĩǀʔã fəŋni tongue ǂheŋ iliŋ tooth totsute !aLkhãH vih water ʔinot ts’aL muku what? hotʃoo ai antq ja who? ho hoo və wide tʃitsaχhu tʃitsaχpho !waHkaL ǁhaLaL ontʃw wind ʔihwokw weHkhẽL sʉvdʉ̃ ʉdʉ̃ wing toɬefw koLbaH asχ wood halʔo pẽLgoL pẽLgeL- mo

toL yesterday tʃəksə you (2sg) am hapu ʃi

137 Table A.18: Toba - Ticuna - Kaingang

Family/ Language Guaykuruan/ Toba Ticuna/ Ticuna Jean/ Kaingang

[ISO 639-3] [tob] [tca] [kgp] ant qagesaq duMɾuLL duFɨM petkɾigŋ arm/hand waq tʃaHHkɨʔHɨM nĩŋe nĩŋndɔ pɛ̃ mḛLL ash halo naʔMHHpaLL̃ mbɾɛj̃ back laq a̰ ʔLLwḛLL pɐ̃nĩ big taHF exe mbɐgŋ bird qojo weLLɾiHH ʃikɾe to bite jenak pɾa bitter tʃim ŋuHHtʃḭʔLLaL fa black naweʔ wḛʔLḛ̃ L̃ ʃɐ blood tagoʔq gɨLL kɨgŋgwedɲ to blow kitʃit jaMkueLL pɐ̃mĩ kɐ̃kɐ bone piʔinek tʃiHHna̰ LLʔãM kuka breast toge miʔLLĩHH nũŋgjɛ fɛ to burn (intr.) ʔedoi jauLɾaLL jauL-

kaLL to carry do jaMŋeH wa mba mɐ̃n child (kin term) jal buʔMH̃ kɾɛ̃ to come wiʔ ŋuM kɐ̃tĩŋ to crush/grind dapaq jaLɲaiʔHHmɨLL ŋgɾɔdn ɲɐ̃ɲ ndɐbm naMkaLL to cry/weep ji ɾɨMdoʔHɨM wɛkɾɛ̃ ñ fɐ̃ aʔMɨM to do/make iʔot iʔet naʔMɨL ɛ kɛ xadn dog pjoq aiLFɾuHH kaʃoɾ to drink jom aLLɨM kɾodn cont.

138 Table A.18 cont. ear tela tʃiMnɨM nĩŋgɾɛŋ̃ to eat keʔe tʃeʔk kiʔi ŋo̰ LL mṵLL tʃiHF- jɛñ ko bɨL egg koʔoueʔ tʃaMɾaMLL̃ ow eye aʔete eHtɨLL kadnɛ̃ to fall qaɲi dʒiM kutɛ̃ far kerom kedom jaʔMHmaM̃ kuwaɾ fire nodek ɨʔLɨM pĩ fish naʎim tʃõʔHFniHH piɾɐ̃ flesh/meat pat maLLtʃiM õHF- nĩ

naLL fly lotaɢaɲi muLɾeHHnM̃ to give ʔan naʔMãM widn to go ke ũH mũ good noʔon meʔLF xɐ hair we jaHHeM ŋɐ̃ɲ kɨki hard qadan taiL taɾ

(s)he/it (3sg) maʒe nɨLFmaʔM fi ti to hear ʔaɢaja mɛŋmɛ̃ ̃ heavy desaʎi jaʔLFũH to hide ɢoxo ɾɨMkṵʔLL jaM- peju

beʔLLkuLL

jaMkṵLL to hit/beat awa naMkua̰ ḭʔLLkaLL pukpɔgŋ tɐ̃ɲ ɾɛñ horn koue tʃaHta̰ ʔLLkuLLɾeM nĩka house nojik maʔ paHtaM ĩʔHF ĩn

I (1sg) ajem tʃaLLmaʔL tʃoLL- idɲ

maʔL cont.

139 Table A.18 cont. in gi guHHneL waHH ki

guʔLxH̃ knee ʎikte ãHHpɨ̰ LLʔɨM jakɾĩ to know ajaten kadɲdɾɔ to laugh laʃi kuHFgɨH ndɨ leaf we fej leg/foot pia japiaʔ kuLtɨM paMɾaLL pɛñ fa liver amek kaLL tɐ̃mɛ̃ long alok ma̰ LL tej louse laɢat naʔLFtʃiLLɾaLL ŋga mouth lap ã̰ LL jɛnkɨ̃ name enaɢat eHHgaLL jiji navel kom ãHHpeLtɨHnaʔLLãM nũŋndidn neck nqosot naʔHãM ndudɲ new dala tɐdɲ tɐ̃ŋ night pe tʃɨHtaMkɨL kutɨ nose mik ɾãHHLL̃ nĩjɛ̃ not saq sa taHmaM tũ old ŋauLL kɔfa ʃi one onolek wɨʔHFiH piɾ rain awot hawot puLkɨH ta red toq dauʔLL kuʃũŋ root paʔa tʃuHHma̰ LãM jɐ̃ɾɛ rope kuHHnaL to run ʔaʃi ɲaMãHtʃiH fũɾũn petɛ̃ wɛɲdwɔ̃ salt toue dʒuMkɨMɾaLL ʃa sand naʔLnɨLLkɨLL ɾɐ̃ɲɾɐ̃j to say aɢat uM kɛ to see lo la met wɛ cont.

140 Table A.18 cont. shade/shadow paɢal tʃiLLpeLLtaM kɐ̃ɲwegŋ fɛndjã skin/hide ʔoq tʃa̰ LLmM̃ fɐɾ small iʔFɾaM ʃĩ smoke malaʔ kaiʔHkeH nĩja soil alwa waiʔMɨHmɨLL ŋga to stand atʃat iLnaLLtʃiH jɛ̃ star waqahɲi ẽʔMtaM woMɾaM- kɾĩŋ

maMkuMɾiLL stone/rock qaʔ nuLFtaM pɐ̃ɾɔ pɔ to suck ʎip naMtʃiH kɐ̃xudn pɐ̃xudn sweet hamae maiʔLkɨHɾaLL ŋgɾɛ̃ tail aʔaɢaʃet ɾeʔFM̃ kɨMɾaH mbɨ to take kona naMnaLF nũkãn nũɲ pednũ thick jaʔHFkɨM thigh teleʔeta peMɾeMmaLL kɾɛ this ʔena ɲaH tagŋ to tie ɲiʔaɢana jaMna̰ ḭʔLL mbɾĩdɲ ʃɛ tɔgfĩn tongue tʃaɢat koHnɨM nũnɛ̃ tooth we pɨLLtaLL jɐ̃ water etaɢat deʔLFaHH ŋgoj what? nege ta̰ ʔLLaHkɨH ndɛ who? nege teʔLLeHH ũ wide qalo taHtʃiMnɨM wind laʔat buLLaHHneHkɨM kɐ̃ka wing awa pḛLLaHHtɨM fɛɾ̃ wood epaq nãĩLF ka yesterday ʃikait ĩLLneʔLL you (2sg) ʔam kuʔLFɨM ɐ̃

141 Table A.19: Pitjantjatjara - Guahibo - Shipibo-Conibo

Family/ Language Australian/ Pitjant- Guajiboan/ Panoan/ Shipibo-

[ISO 639-3] jatjara [pjt] Guahibo [guh] Conibo [shp] ant miŋa amãĩ iβo arm/hand miɳa kobee mãxɨ mɯ pojã ash ipa kuɭku t̻uɳpa iipumãnã tʃiimapo back t̻aɳa humã pɯka big puɭka pinĩhĩ ajai ani bird t̻uɭpu baʐatsui iʂa isa to bite pat̻a sii natɯʂ bitter atsaxa amape apeʐe moka black maɽu hiali tsae tʃɯʂɯ wiso blood milkaɭi t̻ulku hanã himi to blow puu waŋka uxu paja bone tarka pesɨ ʂao breast ipi mĩĩ ʂo to burn (intr.) kampa tahu mɯno to carry kati kapo βo child (kin term) kat̻a uɳʈal xi xɨnã βakɯ to come pit̻an̻ i pata hoti to crush/grind ruŋka tsaki βitaʂ rɯnɯ to cry/weep ula ilun̻ i n βɯ̃ õ to do/make pal̻ a exa a dog papa t̻itut̻a aviʐi otʃiti to drink t̻iki a ʂɯa ear pina mũxu paβiki to eat muŋu x pi haɰɯkia egg ŋampu tobɨ batʃi eye kuɽu itaxu βɯro cont.

142 Table A.19 cont. to fall pat̻aŋaɽa hoo otho poʂo ka paskɯ far parari paʈu taahɨ otʃo fire waɽu tʃii fish antipina duhuai ñ ̃ waka flesh/meat kuka kuka pevii nami fly punpunpa dainã puuluu hoʃinaka to give uŋa rahu mɯni to go ana po ka good pal̻ a xa hakõ hair maŋka mãta nãã βoo hard taiti witu kanti ataha tʃoriʃ

(s)he/it (3sg) paluɽu poñ povaa xua ha to hear kuli nĩka heavy puɭka aʐeve iɰɯ to hide kumpi mãtaja honɯ to hit/beat puŋa b tseko tʃoʃa taʃa tima horn milpin̻ milpin̻ pa mãtʃã house waɭi boo ʂoβo

I (1sg) ŋajulu xañ ɯa in ŋka la a ta ãĩ knee mut̻i raboʂo to know ninti onã to laugh ikariŋa osã leaf n̻ alpi xu pɯi leg/foot t̻aɭpa sii topaa taxu taɯ witaʂ liver alu hapa taka long waɽa piapia nɯ̃ kɯ louse t̻iʈu taʐɨ ia iki mouth t̻aa pekibo kɯ cont.

143 Table A.19 cont. name ini vɨñ hanɯ navel puɭi pokotɯʂɯ no neck liri tɯʂo new kuwarit̻a pɨhanã βɯna night muŋa mẽʐavi jamɯ nose mul̻ a pumũũ re not wija apo ma old iritit̻a peʐuhu pajo rɯkɯ̃ one kut̻u kae ɰɯstiora rain kapi mina emã oi red rituwana tsoobia hoʃĩ root iwiɽi tapõ rope ruupa mã̃ risβi to run wirt̻apaka iʃto salt t̻altu jaho taʃi sand taɭi maʃi to say wat̻a ha a to see n̻ aŋa t his shade/shadow kaɳku wilt̻a βɯi skin/hide miɽi pantu pii boko βitʃi small t̻ukut̻uku t̻apu tsikiʐi hatsi anima smoke puju t̻unaɳ ɨtsɨxɨ kõĩ soil manta atsa iʐa mai to stand ŋaɽa nũũ nina star kililpi tulupu ɥiʃtĩ stone/rock puɭi apu ibo ʃãka makã to suck kuurt̻ana kuunt̻ana tsutsu noʂo tsitsi ojo sweet t̻urat̻a xaika bata tail wipu bosoo tobe hina cont.

144 Table A.19 cont. to take kati pi βo thick ajai itajaɨ kɯʂto rastõ thigh t̻unta bii thaɨ kiʃi this n̻ aŋat̻a mãã nato to tie karpi katonõhõ kɨ keʐe nɯʂa tongue t̻aɭin̻ t̻aɭin̻ pa eebaʐɨ han tooth kat̻it̻i wonõ maka water kapi mina piitaʐaanãẽ mẽnẽ ompaʂ hɯnɯ

mẽʐa what? n̻ aa dee jaɰɯ who? ŋana hipa tsoa wide lipi ajai pɯʂtin naʂβa wind waɭpa howiibo niwɯ wing n̻ alpi pekoʐoɸeto pɯi wood puɳu ŋaɳa hiɥi yesterday muŋart̻i kanĩvi βakiʃ you (2sg) n̻ untu xam̃ mia

145 Table A.20: Yanomamö - Tucano - Warao

Family/ Language Yanomaman/ Tucanoan/ Tucano Warao/ Warao

[ISO 639-3] Yanomamö [guu] [tuo] [wba] ant koje ʃihõ kaʃi meLkaR muha arm/hand imikɨ aLmuH haɾa moho ash nɔLhaH hehuhu back həmapɨ sḛLeHmã ̰ L dai big pata paLhiH paLkaL iɾa bird kiritamɨ miLɾiLkLhɨR̃ domu to bite wə kṵ̃LɾiH bũ bitter koami sLeH̃ aheɾa black iʃĩ sɨLɾiL ana blood iə̃ diLiH hotu to blow hoɾa puLtiR ahakã bone ũ ɔ̰ ̃LaH muhu breast paɾɨkɨ kuLtiHɾɔL ɔ̃Lpe- meho mi HɾɔL to burn (intr.) dokũ to carry mao ɔ̰ LmaH konarũ child (kin term) təə ihiɾu maLk noboto to come a̰ tiH naõ to crush/grind katik patik tḭLaH̃ dasĩ sikarĩ hahĩ to cry/weep ɨk̃ uLtiH onã to do/make ta ʋeLeR nonã dog hiima diLaHjɨL beɾo to drink ko sḭLɾiR̃ hobĩ ear jɨmɨka ɔ̰ LmeH kohoko to eat wa ia tehi ba̰ LaH nahoɾõ egg nate diLeR hĩ cont.

146 Table A.20 cont. eye mamo kaLpeH mu to fall ke bɨLɾɨH nakã far pɾaha jɔLaH emo itõ fire wakə mḛLeL he fish juɾi maɾoha ʋa̰ LiH homabaka flesh/meat dḭLiR toma fly moo hikiri to give hipə ɔ̰ LɔH moã to go hu ʋa̰ LaH naɾũ good totih ãLjuR jakeɾa hair hena koĩ pɔHaL hio hĩo hard bɨLtiH jɔ̃LkɔLaH turu (s)he/it (3sg) pə kɨLɨH kɔLɔH tao to hear tɨ̰ LɔR nokõ heavy hute nɨLkɨH uɾi uɾu to hide hitə nɨLɔR dṵLtiR esotõ to hit/beat ʃə paLaH kẽLeH ahĩ jeweɾẽ tainĩ horn joɾã sa̰ LaH hohi house jahi ʃapono jano ʋḭLiR hanoko

I (1sg) ja jɨ̰ LɨH ine in pɨL eku knee maheko muku to know maLsiH naminã to laugh ikã buLhiR enõ leaf pũLɾiH aɾoko leg/foot mata mamikɨ dɨ̰ LpɔH jẽLkaH ka omu liver amokɨ mahi long ɾap jɔLaH hiwarã louse noma ḭLiL mĩ cont.

147 Table A.20 cont. mouth kahikɨ ɨLseHɾɔL doko name wahã ʋaLmeR wai navel jimotaʃi suLmuHgaL kabu neck oɾahitə ʋaLmɨHtaLhaL dõ new tute ʋḭLmaR ma̰ LmaR hiɾo night titi jaLmiH ima nose hũka pɔ̃LeL hikaɾi not mi pə tiL ʋḛLeL naka old hote bɨLkɨL meLhaH nobo one mahu nḭLk hisa rain maa aLkɔHɾɔL naha red wakə sɔ̰ ̃aH simo root nasiki nɨ̰ LkɔH kahaɾe rope daLaL tao to run ɾəɾə ɔLmaR hakã salt mɔLaH bamu sand maka waha to say ku niLiH dibũ to see ta ḭLyaH̃ mĩ shade/shadow imahana skin/hide si kaLseHɾɔL hoɾo small waĩsipɨ kã̰ L ɔ̰ ̃LɔHmaLhaL sanuka smoke ɔ̰ LmeR hehuku soil maʃita pita dḭLtaH hobahi to stand upɾa nṵLkuR kanamã star puɾiwaɾi jɔ̃LkɔLaR kuɾa stone/rock maa ɨ̰ LtaR hojo to suck wasus usus sḭLsiH uLhuH bahã sweet keteti ḭLpiHtiL diaba cont.

148 Table A.20 cont. tail ʃina pĩLkɔHɾɔL hu to take tə miLiH oã thick supɾe dɔLbɔL tɨLbɨL ukabaka thigh waku ɨLsɔH doɾo this hei a̰ LtiR tama to tie dɨ̰ ɾeR kojã tongue aka jḛLmeHɾɔL hono tooth nakɨ uLpiHkaL hi water aLkɔR ho what? eʃi jḛLeR̃ jaLmɨR bitu jaLmɔR who? weti nɔLaR sina wide seLʋaL beɾẽ wind watoɾi ʋɨ̰ LɾɔR ahaka wing hakosi ʋɨLɨLɾɔR wihi wood juLkɨR dau yesterday wɨjaha kãLɾɨH jaLmiHka̰ L kahe you (2sg) wa mɨ̰ LɨH ihi

149 Table A.21: Awa-Cuaiquer - Mai Brat - Piaroa

Family/ Language Barbacoan/ Awa- Maybrat/ Mai Brat Sálivan/ Piaroa

[ISO 639-3] Cuaiquer [kwi] [ayz] [pid] ant im ii iʔtsotɨx arm/hand tʃɨ̥ tti ʃɨttɨ sail natem tʃuxmuɨx ɨxoʔ ash im box aɲaɾiʔtsaʔ back puntsɨ̥ puns tʃuɾuʔpɛxkex

ɨxkeaʔkex axax big katsa mabox ɾiaʔaʔ ɾixanaʔ ox-

txox bird ʃitʃu ru pxiʔjuɛʔ to bite sula kwa maɸit ɲiʔexu bitter mtax black tɨl saɸe ɨnɨkuaʔ ɨdɨkuaʔ blood ajmbi mes to blow us nɸie pxuɨʔɨ bone tʃɨ̥ mtai iʔwekʔaʔ breast ʃitʃu ʃutʃu tʃitʃu masis kaxtɛteh to burn (intr.) aiz to carry iŋ kwaa nasom nti xenpuɨ child (kin term) paʃpa ku ɨxtɨx to come a ɨxtʃɨ to crush/grind kɨɨ kʔaɾɨ to cry/weep az nawia axɛʔu to do/make ki nno xeʔpuɨ dog kwiʒa mtax ɨʔwɨɾɨx ɛʔwɨɾɨʔ to drink ku kwa nata aʔux ear kail nimara tʃaxax to eat kwa ku nait kuɛʔnɨ cont.

150 Table A.21 cont. egg wiβu ɨʔjaʔjaʔ eye kasu nasu tʃiaɾex to fall najŋ taiz mbtek moaʔtʃe far atti̥ ʃ ɸari ee oxtox fire iiŋ taɸox oxkuʔɾax fish piʃkatu ʃox pɔɨʔtsax flesh/meat ɲɛ nkiwan deʔax fly pak buba naʔnɨex to give kwin mwɨla nee xeʔpuɨ to go ɨ namo tʃɨʔɨ tʃaxtɨ good wat mooɸ aʔdiʔwaʔ adiwaʔaʔ hair aʃ mawian tʃuʔwoʔtsex tuʔ-

wotsex hard sam matak iʔjaʔaʔ

(s)he/it (3sg) us ait au jaxux tʃuʔ to hear mwɨ siende nari heavy us ɸanes samjox xamakaʔaʔ to hide kula pwin nmut nmut daxaʔwaʔtʃi to hit/beat pjan kuz nai kuaxkuaʔpuɨ horn pul mankaron axaʔtsaʔɾix house jal samu oxuʔtsoʔdex

I (1sg) na dzo txɨʔ in ta mam xuaʔaʔ knee wakpul nimna ɨxmuɾexkiax to know pjan nxar exuku to laugh iʃan nasah leaf kɨ̥ mata iʔtsoxɛʔ leg/foot pimpul mittɨ naa tʃixepuɨx tɨxtaɾe-

maʔ cont.

151 Table A.21 cont. liver ɲantunul mwan long akki̥ ʃ mara manaɸ maa oxtoaʔa louse mũ xate kuɨʔɨx mouth pit nasox tsɛʔ aʔ name mun navel pamtʃin xrobit tʃuɾawaʔ neck kwi’ srau uɾuʔpɛkʔe new maas tna xaɾexnɛʔ night amta mti jɔʔdɔʔ nose kimpuh nnaiɸ tʃixiʔɲux not ʃi ɸe old ɨlaβa nabi mabi buɨʔoʔatsaʔ one maza sau ɲantɨx rain alu oom axijaʔ red kwandam mkek root muxtɨt mtis uxukɛxkex rope tɨt too aʔnutsox to run nxox xoʔpɛʔtsi salt ɲɛm xa sand kɨ̥ su sɲik ɾeʔma to say kiʒ nawe peɨx to see iʒ nmat taoʔpuɨ shade/shadow skin/hide aja marak ixetaʔ ixetax small ajngi maku kintʃaʔ smoke iʃ wasi soil pil tabam to stand kun pana nros kxɨʔɨ star kɨɨma sken tsiɾikoʔoʔaʔ cont.

152 Table A.21 cont. stone/rock uk ɸra ɨʔdokɨx to suck tuk tmos tnos nmot ɾaxadɨʔɨ ɾaʔadɨ sweet msu tʃaʔnaʔwaʔ tail mwɨɾa mɨtɨ miɸra ɨʔwɨɾɨʔpaʔ to take mwia tʃuaʔdɨ eʔmuɨ thick katsa xonto baɾempax thigh ɲamtʃɨh naɸamu tʃixaʔ this ana piʔɲexkɨx to tie pal ip kuht nake nkxamit nɨʔkɨ tongue jankwixia xrea iʔnex tooth sula mbait tʃaxkux water kwazi aja axijax what? ʃi bawia daxeʔ who? mwɨn awia dɨxex wide aitʃ mase wind ingwa ɸos ɨʔtsopxax wing iltʃap maim ɨxaʔbɛʔ wood ara dauʔtsoʔdaʔ yesterday andʒik iis you (2sg) nu ɲo kuʔ

153 Table A.22: Amanab - Choctaw - Cherokee

Family/ Language Border/ Amanab Muskogean/ Iroquoian/ Chero-

[ISO 639-3] [amn] Choctaw [cho] kee [chr] ant homp laugwe ʃukani tosətaɬi arm/hand niŋa ibbak ʃakbaʔ khanokeni ɨwojeni ash sumof susobug ʃinap kostɨ back masig naliʔ kasohi big kwania tʃito ekhwa bird tuwafig hoʃiʔ thsiskhwa to bite neg kopooli skala bitter miakla takbah ɨjəsti black yimilig losa kən̥ ake blood taf issiʃ issis kikʌ to blow fwiafe maɬi apeli pũfa tatəs bone kig foniʔ koɬa breast tut ipĩʃik kanethsi to burn (intr.) tuhlu lowa kalejəsə to carry uku ia awitʌti child (kin term) luwal oʃiʔ allaʔ ajoli to come blo ahii kalɨthsə to crush/grind pigiasko kaɬali litoɬi stos to cry/weep ufwe jaaja thlos to do/make fe mihtʃi jmbj n̥ atə gotlə dog malaŋ ofiʔ kithɬi to drink ne iʃko tithas ear aŋug haksobiʃ kaleni to eat ne ĩpa apa kis egg suwig akãkoʃiʔ ɨwethse eye nof niʃkin kahthooli cont.

154 Table A.22 cont. to fall pe ittola ɨlosə far membi atapa fehna hopaki inə fire suwi lowak athsilə fish sawag naniʔ athsata flesh/meat nihig nipiʔ haknip hawija ɨkwedaɬi fly amof tʃukani ʃũʃi to give fai a aa ima atane to go ika ija nekes good suialag atʃokma osta hair lapog hiʃi pãʃi ɨsthikə hard nokig kamassa kallo ɬim- astaja

impa

(s)he/it (3sg) ehe naski to hear hig haklo tə heavy kinig wiiki kəketə to hide kumohlai lohmi loma tiskalo to hit/beat tige isso boo tatəs datɬohis horn aniŋuog lapiʃ ɨjona house lala tʃokkaʔ kaɬithsotə

I (1sg) ka anoʔ aja in saiye aa ʔi knee mugukug tininikeni to know anwenafe khaan thə to laugh titge joppa jethsas leaf figeg hiʃi ɨkalokə leg/foot muŋugag hanaaliʔ ijjiʔ nəske lasidena liver gufwag salakhaʔ ɨwela long kwieg falaia hofaloha kanəhi louse ku issap thina cont.

155 Table A.22 cont. mouth maŋig itakhaʔ aholi name unehla hotʃifo hohtʃifo thoʔə navel usumu hatãbiʃ neck wuneglig ikkõlaʔ akilakeni new yimieg himmona inla ithse night sinaini ninak sənoji nose bosog habiʃkoʔ ibiiʃakniʔ kajəsoli not mas ok o kiijo thɬa old watiaug sipokni akajə one muŋu atʃaffa sakhu rain bu umba ka red tafnaiig homma kikake root bagag akiʃtala unastethɬə rope we aseta ponokallo stejitə to run kikinika baliili salt sol hapi ama sand eŋig ʃinuk noja to say sihi mija aatʃi hineki to see naŋu pĩsa ko shade/shadow sugwabug hoʃonti tawatiskə skin/hide lofug hakʃop kaneka small weti iskitiini jooɬi smoke suwiel ʃoboli soil bite lukfi lusa jakni katohi to stand go hikĩja katokə star bagnofut fitʃik nokhsi stone/rock fun taliʔ nəja to suck asanktata ʃõka sweet tʃãpoli kanas cont.

156 Table A.22 cont. tail efleg halĩbis hasĩbis kanitatə to take fla iʃi ʃaa kis thick hofulug toluski ɨhaketə thigh bigisig hattip this hwa ilappa hiʔa to tie kike ʃekonobli taktʃi khane tongue meleg issõlaʃ kanək tooth gog notiʔ kanətokə water bu okaʔ ama what? maŋo nãta kato who? amban kata kako wide wawag awatta ajathena wind efef fiopa maɬi kanoləs wing aŋlag aɬi sanatʃi kojate wood aga ittiʔ ata yesterday amu pilaaʃaaʃ sahi you (2sg) ne tʃiʃnoʔ nihi

157 Appendix B Lexical sources and transcription method of each language

All the sources used are from the 20th-21st century.

Table B.1: Lexical sources and transcription method of each language.

Language Source Spanish Lexicon: Personal knowledge. Transcription: [42] Mandarin Chinese Lexicon: Personal knowledge. Transcription: [43] Yoruba [44] Egyptian Spoken Arabic [45], p. c. with a speaker Javanese [46, 47] Telugu Lexicon: [48, 49, 50]. Transcription: [51] Turkish Lexicon: [52]. Transcription: [53] Japanese Lexicon: Personal knowledge. Transcription: [54] Vietnamese Lexicon: Personal knowledge. Transcription: [55] Thai Lexicon: Personal knowledge. Transcription: [56] Korean Personal knowledge (I am a native speaker) Dholuo [57], [58], [59]

cont.

158 Table B.1 cont.

Hungarian Lexicon: [60], consultation of an online dictionary. Transcription: [61] Chuanqiandian Cluster Miao [62, 63, 64] South Bolivian Quechua [65, 66, 67] Peripheral Mongolian [68, 69]. Note: Chakhar dialect. Kabardian Lexicon: [70, 71, 72]. Transcription: [73] K’iche’ [74, 75] Paraguayan Guarani [76] Georgian [77, 78], p. c. with a native speaker Enga P. c. with Adam Boyd (Enga Bible Translation Program) Eastern Huasteca Nahuatl P. c. with John Sullivan (Uniwersytet Warsza- wski and Universidad Autónoma de Zacatecas) Central Aymara [79, 80] Meqzuital Otomi [81, 82] Wayuu [83] Basque [84] Ngäbere [85] Highland Totonac [86] Khoekhoe [87] Galela Lexicon: P. c. with a native speaker. Transcrip- tion: [88] Mapudungun [89] Western Highland Purepecha [90, 91], p. c. with Claudine Chamoreau (CNRS) Woods Cree [92], p. c. with Arok Wolvengrey (First Nations University of Canada) cont.

159 Table B.1 cont.

Navajo [93] Highland Popoluca [94], p. c. with Lynda J. de Jong Boudreault Ambulas [95], p. c. with Richard Scaglion (University of Pittsburgh) Mískito [96, 97] Shuar [98, 99] Northern Emberá [100] Bukiyip Unpublished field notes of Lise M. Dobrin (Uni- versity of Virginia) Northwestern !Kung [101], p. c. with Bonny Sands (Northern Ari- zona University) Greenlandic [102], p. c. with Minik Jeremiassen (Language Secretariat of Greenland) Burushaski [103, 104]. Note: Yasin dialect. Sentani [105] Terei [106] Macushi [107] Ap Ma P. c. with Jill Riepe, p. c. with Jesse and Karie Pryor (Outreach International of Papua New Guinea) Páez [108, 109]. Transcription: [110] Wichí Lhamtés Vejoz [111, 112] Sandawe [113, 114]. Note: Tones not transcribed in [114] Xibe [115, 116, 117] Toba [118, 119, 120] Ticuna [121] Kaingang [122] cont.

160 Table B.1 cont.

Pitjantjatjara [123, 124], p. c. with Cliff Goddard (Griffith University) Guahibo [125], p. c. with Francesc Queixalós (CNRS) Shipibo-Conibo Lexicon: [126, 127]. Transcription: [128] Yanomamö [129] Tucano [130] Warao Lexicon: [131, 132]. Transcription: [133]. Note: Only the morpheme for ‘hard’ retrieved from [132], an online dictionary. Nasal diacritics are not transcribed in [132]. Awa-Cuaiquer [134] Mai Brat [135] Piaroa [136] Amanab P. c. with Andy Minch (SIL) Choctaw [137, 138] Cherokee [139, 140]

161