[now see also:Ascent of Intelligence and How Children acquire Language]
Language and Evolution: Books, Presentations and Papers
Motor Theory of Language
Back to
Chapter I Hypothesis:Phonological/Semantic Equivalence
Chapter II Parallelism of Speech and Gesture
Chapter III Speech-sound and Gesture Elements
Chapter IV Verification: Relation of Sound and Meaning
EVIDENCE FROM OTHER LANGUAGES
Expressivism in foreign languages
Resemblances between remote languages
Uniformities in demonstratives and pronouns
Uniformities in naming colours
Sound/meaning relation in Basque
CONCLUSION
REFERENCES
Chapter IV presented evidence for the existence of a systematic relation between sound and meaning in the vocabulary of the English language; references to other languages were limited and incidental. The conclusion reached was that there is good evidence of a systematic relation of sound and meaning in English and that the expressiveness of words in English, their appropriateness for their meaning, is linked to and can be explained by this. Nevertheless, it remains possible though not very probable that the apparent relations between sound and meaning in English and the apparent expressiveness of many words may be a purely English phenomenon. However, if in English the sound/meaning relation is not cultural or conventional but natural, there is no reason why foreign languages should not provide evidence to support this in the words they employ. Beyond this, if more languages than English display internally systematic relations of sound and meaning, the question arises how far the specific system for relating sound-elements and elements of meaning is something common to many languages, or possibly the universal underlying foundation for language generally; the enquiry then becomes how far a regular relation can be established between the systematic pattern of sound/meaning in English and sound/ meaning patterns in other languages.
The propositions considered in this Chapter are the following:
A Expressivism - that is the phenomenon that words are felt to be appropriate to their meanings - is as well-established for foreign languages as it is for English, particularly for French, Spanish and German. There is evidence of expressivism also in remoter languages;
B The expressiveness of words in English can be appreciated by speakers of foreign languages. Even though the patterns of sound may be unfamiliar to the foreign language-speakers and the words do not for them form part of any traditional sound/ meaning system, the expressive force is still conveyed;
C In foreign languages, apart from expressiveness in the sense of the felt appropriateness of particular words, there is evidence of sound/meaning groupings of words formed in a similar way to the groupings in English discussed in the previous Chapter i.e collections of words where the sound and meaning appear to vary together in a systematic way;
D For a considerable number of high-frequency words, naming familiar objects and actions, where the likelihood of borrowing between languages is small, one can find a 'mass-language effect', that is resemblance of sound and meaning which spread across many languages, related and unrelated, including languages geographically and in terms of language family extremely remote from each other. The resemblances go beyond anything that can plausibly be explained as chance or coincidence;
E This is particularly the case for some simple basic words where the meaning is sharply defined, notably for pronouns and demonstratives, for the names of the major colours and for words used to express some family relationships.
Some evidence on this was incidentally presented in Chapter IV. The object in this section is to look more closely at the position for a few foreign languages. For the French language, the most useful discussion is in Sauvageot's Portrait du Vocabulaire Français (1964). Though he is a strong adherent of the Saussurean school, he diverges from de Saussure in stressing the great importance attaching to onomatopoeic and expressive words in the French language (de Saussure had suggested that onomatopoeia was of marginal importance). According to Sauvageot, if one listens to the spoken French language rather than concentrating on the written language, one soon discovers that, even among educated people, the language "fourmille d'onomatopées les plus variées". Apart from onomatopoeia "L'expressivité générale de l'énonciation française s'est affirmée. Le français s'emplit de néologismes expressifs.. Les Français ne parlent donc pas plus 'abstrait' que les Anglais, les Allemands, les Russes &c. On a même le sentiment qu'ils seraient plus portés que leurs autres contemporains à confier l'expression de leur pensée quotidienne à des procédés d'allure onomatopéique".
Much the same can be said about Spanish. Garcia de Diego says: "We have to start from the principle that all words in varying degrees have a sensory or emotive symbolic value. In some words this value is obvious and commonly recognised, for example in words which refer to an action which is accompanied by an audible sound. This is the well-known symbolism of SILBAR (whizz) or ZUMBAR (buzz). All peoples at all periods have recognised that certain words evoke particular sensations or ideas." Spanish literature is rich in the use of expressive words and in Spanish the expressiveness of sounds is a well-established fact.
In the German language, Jespersen referred to the work done by other writers to demonstrate the extent of expressive sound, for example Hilmer's study of apparent links between particular sounds in German words and particular shapes and appearances of things - which extended to 70 pages of word-lists of expressive words in German with their English equivalents. Humboldt quoted as obvious examples of sound symbolism words like WEHEN WOLKE WIRREN WUNSCH where the vacillating, wavering motion is expressed through the letter W. Firth discussed the wide range of expressive words in Dutch and the Scandinavian languages. He noted in Norwegian about 80 pejorative words beginning with SL and a similar number in Swedish. He also listed 25 'nasal' words in Norwegian beginning with SN (parallel to English words like SNIFF SNIGGER &c); he found some 70 words beginning with FL belonging to a group like that of the English words FLECK FLICK FLICKER. Initial KL appears for 80 "clumsy, cloggy, ungainly, sticky" words. In Dutch he gave examples of a large number of expressive words including as a choice example PLIMPPLAMPPLETTEREN meaning 'to play ducks and drakes' that is throwing a flat stone to make it skip as many times as possible across the surface of water.
Thus the expressive force of words - the feeling that they are peculiarly appropriate in many instances to their meanings - is well-established for a good number of languages. The languages mentioned belong to the Indo-European group. There has been less discussion of sound symbolism, word-expressiveness, for non-Indo-European languages. There is no reason to think that it is any less important for Semitic and African languages, for the Polynesian languages referred to by Paget and Rea or indeed for the aboriginal Australian and Amerindian languages. In Les Langues du Monde, Manchu for example is said to have been absolutely full of imitative formations. Hörmann quotes examples from African languages of sound symbolism: in Ewe high-tone words indicate small things and low-tone words large things; in certain Sudanese languages high-tone words are used to express long distances or high speed and low-tone words to express proximity and slowness. However, a complete examination of expressiveness in remoter languages is best done by native speakers of the languages since they are able to answer the essential question: how appropriate to the meaning is the sound of the word felt to be, against the background of knowledge of the whole range of sounds and meanings of words found in current speech in any particular language.
Nevertheless expressiveness of words in one language can be appreciated by speakers of another. Recognition of expressiveness is not necessarily limited to native-speakers of the language. Sauvageot demonstrates this for French-speakers. In modern times there has been extensive borrowing into French of English (and American) words. He observes that some words have been introduced not because French lacks a suitable word of its own for a particular meaning but because the expressive force of the English word has led to its being used in French in preference to the French word. Onomatopoeic words are borrowed between languages in the same way as other words. "So in modern French we have borrowed the BANG of the English. Our engineers speak about the THUMP or the noise produced by the tyres of an automobile in contact with the surface of the highway." He speaks somewhat disapprovingly about "quelques intéllectuels français, peu initiés à la linguistique qui emboîtent le pas en se pâmant d'admiration devant la force suggestive de certains vocables anglais tels que THUMP BANG RUSH SPLASH FLASH CRASH &c". This is interesting because words like THUMP BANG SPLASH are of a phonetic form unfamiliar in French. Any expressive force can hardly derive from 'phonaesthetic habits' established by spreading association within French vocabulary. The appeal of such unfamiliar but expressive words seems natural rather than in any sense cultural or conventional.
It is of more interest perhaps to look at the extent of expressiveness in a remoter language outside the Indo-European family, namely Malayan. It is possible in examining Malay vocabulary to appreciate quite readily (as Sauvageot can for expressive English words) the peculiar appropriateness of a good number of words to their meanings. The following list of words in Malayan has been drawn up to demonstrate this. The Malayan words are presented first and then the corresponding English words in the same order It is interesting to study the Malayan words first to judge their expressiveness and possible meaning before looking at the English equivalents.
EXPRESSIVE MALAYAN WORDS
BERLETAR KELEPOK KEKEK-KEKEK PUNCHA MEREMAH-REMAH
EMBEK KEPITAN CHEKEK PUKULAN MEMUTAR-MUTAR
BELAHAK KETAM GELOGO PUCHOK BERTATEH-TATEH
PELOK MENGOGOK PAKU
BERPUPUT MELEKAT MENYALOK PATOK TIPIS
PUPUTAN KEPITING MENJERIT PILEH TEBOL
BUAK SA-RUMPUN MENGATAKAN PERET
BUEH BERKUMPUL MENGUAP PETEK BESAR
MENGUAPGUAP PIDI KECHIL
BUAH DADA GENGGAMAN NGAP-NGAP
BUNCHIT MENGGILING PAPAN
GEMBONG TANGKAPAN MERUNGUT PANGGOK
GELEMBONG MENGHANCHORKAN MERENGUS PANCHANG
MEMBUKA KERKAH BERSUNGUT
BISUL KUKU GANAS
BOLA KEKOK MURAM
BATIL GERANG
(BOTOL) RIANG
KERIANGAM
ENGLISH EQUIVALENTS
BABBLE CLAP GIGGLE TAB CRUMBLE
BLEAT CLIP GAG PUNCH TWIDDLE
BELCH CLAMP GULP BUD TOTTER
CLASP GURGLE PEG
BLOW CLING YAP PECK THIN
PUFF CRAB YELL PICK THICK
BUBBLE CLUMP TELL PRICK
FOAM CLUSTER YAWN PLUCK BIG
GAPE PITCH LITTLE
BREAST GRIP GASP
BULGE GRIND PLANK
BLOATED GRASP GRUNT PROP
BLADDER CRUSH GRUFF POLE
BURST CRUNCH GRUMBLE
BOIL CLAW GRIM
BALL CLUMSY GLOOMY
BOWL GLAD
(BOTTLE) GAY
GLEE
EXPRESSIVE MALAYAN WORDS (Continued)
BAWAH KERTIK MELIHAT PALAM LOMPAT
BUNTUT MERETAKKAN MEMANDANG PUNG LOMPOK
BENTOK MENGETIK PANDANG CHELAM-CHELUM LIMPA
BERGABONG MENDETIK KILAP GUNTAR LIHAT
BUNGKUSAN GELITEK KILAT PALONG LESTARI
BENGKOK DETING GILANG-GEMILANG GELEPONG LENTING
BONGKOK KEJANG MENGILAP GELETING LENGAH
BUNGA KILAP GELETOK LEBAS
PUTAR KELIP DENTUM LAKSANA
BERPUTAR MONGMONG LACHUT
MEMUTAR-MUTAR
SAMA
NAMA
NOT
YA
WAI
TUTUP
TUTOR
TUNGKAP
TUB-TUB
ENGLISH EQUIVALENTS
BOTTOM CLICK LOOK PLUG JUMP
BUTT CRACK GLANCE PLOP LUMP
BEND TAP STARE THUMP LIVER
BUNCH TICK GLIMPSE THUNDER LOOK AT
BUNDLE TICKLE GLINT POOL LASTING
CROOKED TWANG GLITTER PLOSH LEAPING
HUMPED TWITCH GLOSSY TINKLE LINGER
BLOSSOM FLASH TOCK LASH
TURN FLICKER BANG LIKE
TWIRL GONG WHIP
TWIDDLE
SAME
NAME
KNOT
YES
WATER
SHUT UP
UTTER
TONGUETIED
SUDDENLY
From examination of the words in the list, it will be seen that the Malayan language forms imitative and expressive words through the use of sound-sequences similar to those familiar in English. The comparison between the group of words in Malayan beginning KELEPOK and the equivalent English group beginning CLAP is especially interesting since the Malayan group seems to show a parallel change in sound and meaning in the same way as was discussed for the group of English words in Chapter IV. The differences from English in the sounds used for expressive purposes and the way sound/meaning groups are constituted are also of interest. For example, Malayan links the words for FLASH FLICKER with those for GLOSSY GLITTER (KELAP KELIP MENGILAP GILANG-GEMILANG) - though in English there is a relation of meaning between GLOSS and FLOSS, and GLITTER and FLICKER have a considerable resemblance to each other.
Some of the Malayan words seem so obviously expressive through their sound that, given the general context, one might well be able to guess what they mean without using a dictionary e.g. KEKEK-KEKEK for GIGGLE, GELITEK for TICKLE, BERPUPUT for BLOW with PUPUTAN for PUFF, KERTIK for CLICK, GELOGOK for GULP, NGAP-NGAP for GASP, KILAP for FLASH, PETEK for PLUCK, PUNG for PLOP, GUNTAR for THUNDER, MONGMONG for GONG, TIPIS/TEBOL for THIN/THICK, LOMPAT for JUMP - and there are also some near-identities as SAMA NAMA NOT YA for SAME NAME KNOT YES.
For the existence of groupings of expressive words in other languages the list just considered for Malay is some evidence. There is obviously material that could be presented on similar lines for more familiar languages. So there is some parallelism between words in French beginning with CL CR GL GR and those beginning with the same initial letters in English - though words starting GR in French may begin CR or CL in English i.e. the resemblance of the forms is a broader one. Some examples are:
GRIFFE CLAW CLAQUER CLAP GRAPPE CLUSTER CLAMEAU CLAMP GRUMELER CLOT CLAPPER CLICK (with the tongue) GRIMPER CLIMB CLAQUER CLATTER AGRAFER CLASP CLIC CLICK (noise) GRIPPER CLUTCH CLIVER CLEAVE GRAPPINER GRAPPLE CLAC CRACK GRUGER CRUNCH CRISPER CLENCH GRUE CRANE CRAMPONNER CLAMP GROUILLER CRAWL(with) CRAQUER CRACK GRINC GRATE CREPIR CRIMP GROGNER GRUMBLE CREPITER CRACKLE GRONDER GROWL CROQUER CRUNCH GRIGNER CRINKLE CROSSE CROOK GROLLER GROUSE CROULER CRUMBLE
Though for the most part no etymological relation has been established between the French and the English words, the cross-linkings between the two languages in their use of a relatively limited number of sounds and sound-sequences to express a similar range of meanings is striking.
The following list shows for 23 languages (including languages widely separated both geographically and in terms of language families) words used for some 40 objects, actions, adjectives and grammatical elements. The objects and actions chosen are those which are well-defined and familiar, the words being among the most-frequently used in ordinary speech 23 of the words relate to parts of the body, there are 9 verbs referring to common simple actions, together with basic nouns like AIR EARTH FIRE WATER &c. This kind of examination could be extended to cover a much larger number of words and a much larger number of languages. However, the less basic the words considered, the greater the chances of borrowing between languages and the less certain any conclusions drawn on the basis of the degree of resemblance observed. Extension of the number of languages would have increased the number drawn from language-groups already represented; addition of more languages would have not added enough to the interest and validity of the results to justify the excessive length to which discussion in this section would inevitably have had to develop.
In the Table which follows the collection of words in the 23 languages (including English) is arranged in alphabetical order as follows:
Arabic Korean
Basque Latin
Chinese Lozi
Finnish Malay
French Samoan
German Spanish
Greek Swahili
Hebrew Telegu
Hungarian Turkish
Italian Venda
Japanese Zulu
In terms of language-families, seven of the languages are from the Indo-European group (Greek, Latin, English, German, French, Spanish, Italian). This has been strongly represented partly because these languages are more familiar and it is instructive to see the variations as well as the similarities between the words they use for familiar objects and ideas, partly because resemblances in remote languages may be to forms of words which appear in some Indo-European languages and not in others. Of the other languages, four are from the African group (Lozi, Swahili, Venda, Zulu) and are more or less closely related to each other, there are the two main Semitic languages Arabic and Hebrew, the two Finno-Ugrian languages Finnish and Hungarian, two languages from the Malay-Polynesian group Malayan and Samoan, and the remainder are clearly isolated languages such as Basque, Chinese, Japanese, Korean or single representatives of other language families such as Telegu (the Dravidian languages) and Turkish (the Altaic languages). Perhaps the main gap is the absence of words from Amerindian languages. There is a bewildering variety of these but, as some check, in the discussion of the words for each separate object or action some reference is made to parallels and resemblances in the Amerindian languages.
In the Table for some words the translation into a few languages is missing either because the language has no separate word for the idea or because the dictionaries available give no translation (particularly so in the case of Lozi). Where a word is missing, a gap is left so that the order of the languages is otherwise preserved.
COMPARATIVE WORD-TABLE ANKLE KNEE KAB POKSONGABYO BATTA MURUP ACHIRAIN TALUS BELHAUN GENU HUA4 NGONGO NA LIUTU HSIAOT'UI LINWELE NILKKA PERGELANGAN KAKI POLVI LUTUT CHEVILLE TAPUVAE GENOU TULIVAE FUSSKNOCHEL ASTRAGALO KNIE RODILLA SPHURON KIFUNDO GONU GOTI YATED CHILAMANDA BEREKH MOKALU BOKA AYAK BILEGI TERD DIZ CAVIGLIA KHOKHOLA GINOCCHIO NGONA ASIKUBI IQAKALA HIZA IDOLO HIP/THIGH ELBOW KHASR NOPJOKTARI KU PALKUMCHI HANKA COXENDRIX UKONDA CUBITUS K'UAN1 - CHOU LANKA PANGGUL KYYNARPAA SIKU HANCHE - COUDE TULILIMA HUFTE CADERA ELLBOGEN CODO GOMPHOS - OLENE YAREKH TODA MARPEK MOCHEYI CSIPO KALCHA KONYOK DIRSEK COSCIA/ANCA TSHIRUMBI GOMITO KOSHI ITHANGA HIZI SHOULDER NECK KITF OKAE 'UNK MOK SORBALDA HUMERUS LEPO COLLUM CHlEN LIHETA CHlNGPU MULALA OLKAPAA BAHU KAULA TENGKOK EPAULE TAU'AU COU UA SCHULTER HOMBRO HALS CUELLO OMOS BEGA AUKHEN SHINGO KATEF BHUJAMU TSAVAR MEDA VALL OMUZ NYAK BOYUN SPALLA SHANDA COLLO MUTSINGA KATA IHLOMBE KUBI UMQALA NOSE HAND MINKHAR KO YADD SON SUDUR NASUS ESKU MANUS P12 NGO SHOU LIZOHO NENA HIDONG KASI TANGAN NEZ ISU MAIN LIMA NASE NARIZ HAND MANO RHINA PUA KHEIR MKONO AF MUKKU YAD CHEYI ORR BURUN KEZ EL NASO NINGO MANO TSHANDA HANA IMPUMULO TE ISANDLA COMPARATIVE WORD-TABLE (contd) EYE EAR 'AIN NUN UDU KWI BEGI OCULUS BEHARRI AUREM YEN3 LIITO ERH3 ZEBE SILMA MATA KORVA TELINGA YEUX MATA OREILLE TALIGA AUGE OJO OHR OREJA OMMA JICHO OTA SIKIO AYIN KANNU OZEN CHEVI SZEM GOZ FUL KULAK OCCHIO MATO ORECCHIO NDEHWE ME ISO MIMI INDLEBE MOUTH LIP TUMM IP SHIFFI IPSUL AHO OREM EZPAIN LABRUM K'0U3 HANU CH'UN MULOMO SUU MULUT HUULI BIBIR BOUCHE GUTU LEVRE LAU MUND BOCA LIPPE LABIO STOMA KINYWA KHEILOS MDOMO PEH NORU SAFAH PEDAVI SZAJ AGIZ AJAK DUDAK BOCCA MULOMO LABBRO MULOMO KUTI UMLOMO BIRU UDEBE TONGUE TOOTH LISAN HYO SINN I MIHI LINGUA HORTZ DENTEM SHE2 LILIMI CHIH LINO KIELI LIDAH HAMMAS GIGI LANGUE LAULAUFAIVA DENT NIFO ZUNGE LENGUA ZAHN DIENTE GLOSSA ULIMI ODONTA JINO LASHON NALUKA SHEN PALLU NYELV DIL FOG DISH LINGUA LULIMI DENTE LINO SITA ULIMI HA IZINYO CHIN/JAW CHEEK DAKN TOK KHADD BYAM MATRAILA MAXILLA MATHEL GENA K'02 MUTAHALI CHIA LEUKA DAGU POSKI PIPI MACHOIR AUVAE JOUE ALAFAU KINN QUIJADA BACKE CARRILLO GNATHOS TAYA PAREIA SHAVU SANTER GADDAMU LEHIY CHEMPA ALL CHENE ORCA YANAK MASCELLA LUTAHA GUANCIA AGO UMLATHE HOO COMPARATIVE WORD-TABLE (contd) HEAD FACE RAS MORI WISH OLGUL BURU CAPUT AURPEGI FACIES T'OU MIENMAO PAA KEPALA MUOTO MUKA TETE ULU FIGURE MATA KOPF CABEZA GESICHT CARA KEPHALE KICHWA PROSOPON USO ROSH TALA PANIYM MUKHAMU FEJ BASH ARC YUZ CAPO FACCIA ATAMA MUKAU VOICE HEART SAUT(HISS) MOKSORI KALB SIMJANG BOZ VOCEM BIHOTZ CORDEM SHENGYIN HSIN AANI SUARA SYDAN JANTONG VOIX LEO COEUR FATU STIMME VOZ HERZ CORAZON PHONE SAUTI KARDIA MOYO KOL KANTHADHVANI LEV HRDAYAMU HANG SEG SZIV KALP VOCE CUORE KOE KOKORO BODY BREAST BADAN MOM BIZZ CHOT GORPUTZ CORPUS TTITTI MAMMA SHEN1TI3 MUBILI HSIUNG LIZWELE RUUMIS-VARTALO BADAN POVI SUSU CORPS TINO SEIN FATAFATA KORPER CUERPO BRUST SENO SOMA MWILI MADZOS KIFUA GUF ODALU DAD PETTU TEST VIRJUT MELL MEME CORPO TSHITUMBU PETTO MADAMU KARADA CHICHI IBELE THROAT ZLA'IM MOK LEPHO GUTTUR YENH0U KURKHU KERONGKONGAN GORGE FA A'l GURGEL GARGANTA PHARUGX KOO GAHON GONTU TOROKX BOGAZ GOLA NODO COMPARATIVE WORD-TABLE (contd) CUT HIT KATA PEGI DARAB MATCHIDA PIKATU CAEDO JO TUNDO K01 PUMA CHICHUNG NATA LEIKATA MEMOTONG LIPDA MEMUKUL COUPER TIPI FRAPPER TA SCHNEIDEN CORTAR SCHLAGEN PEGAR KEIRO KATA TUPTEIN PIGA KARAT KOYU HIKAH KOTTU VAG KESME UT VURMSIK TAGLIARE UTUMULA COLPIRE HIRWA KIRU SIKA UTU BETA GO GIVE RAN KADA 'ADA CH UDA JOAN VADO EMAN DARE HSING YA CHI FA KULKEA PERGI ANTAA MEMBERI ALLER ALU DONNER FOA'I GEHEN IR GEBEN DAR KHOREIN KWENDA DIDOMI PA HALAKH VELLU NATAN ICHCHU MENJEN GITMEK AD ALMAK ANDARE IYAYI DARE UFHA IKU HAMBA KURE NIKA RUB SPIT RAK PIBIDA TAFF CHIMPAETDA TORRATU TERO THU SPUO MOTSA PIKITA T'OT'U KWA MATI HANGATA MENGGOSOK SYLKEA BERLUDAH FROTTER MILl CRACHER ANU REIBEN ESTREGAR SPUCKEN ESCUPIR TRIBEIN KUSUGUA PTUEIN TEMA MATE HIKEKH PARUGETTU YARAK UMMU DORZSOL SURTMEK KOP TUKURMEK STROFINARE UKOROPA SPUTARE UPFA KOSURU HLIKIHLA TSUBA AMATHE PRICK LICK SHAKK JIRUDA LIHIS SISTATU PUNGO MILIKATU LINGO CHlENTZU TABA SHIH LAZWA PISTAA CHUCHOK NUOLLA MENJILAT PIQUER TUI LECHER SALOSALO STECHEN PICAR LECKEN LAMER KENTEIN CHOMA LIKHMADZO LAMBA DAKAR PODUCHU LIKEK NAKU SZUR SOKMET NYAL YALAMAK PUNGERE MUPFA LECCARE NANZWA SASH HLABA NAMERU KHOTHA COMPARATIVE WORD-TABLE (contd) AIR/WIND WATER ARYAH KUNGI MWAI MUL AIRE AER UR AQUA FENG MOYA SHUI MEZI ILMA ANGIN VESI AYER AIR EA EAU VAI LUFT AIRE WASSER AGUA AER HEWA HUDOR MAJI AVIYR GALl MAYIM NILLU LEVEGO ICHI VIZ SU ARIA UWAINDA ACQUA MADI KUUKI UMOYA MIZU AMANZI FIRE EARTH NAR PUL ARD CHIGU SU IGNIS LUR TERRA HUO MULILO TUTI MUBU TULI API MAA BUMI FEU AFE TERRE ELE FEUER FUEGO ERDE TIERRA PUR MOTO GE NEHI ESH NIPPU ODAMAH BHUMI TUZ ATES TUTI TOPRAK FUOCO MULILO TERRA MAVU HI UMLILO RIKU UMHLABA WAVE CRAB MAUJ SARATAN KE UHAIN UNDA KARAMARRI CANCER PO LINDINDA HSIEH NKALA LAINE OMBAK RAPUELAIN KEPITING ONDE GALU CRABE PA'A WOGE OLA KRABBE CAMBARO KUMA WIMBI KARKINOS KAA GAL ALA SARTAN KAPPU HULLAM DALGA TENGERIRAK YENGEC ONDA LIGAMBEIO GRANCHIO DAMBATSHEKWA NAMI IGAGASI KANI INKALA DEAD MAIYIT CHUKUN HERIOTZE MORTUUS SSUTE MUSHU KUOLLUT MATI MORT MATE TOT MUERTO NEKROS MUFU MET CHACHCHINA HALOTT OLU MORTO MUFU SINDE FILEYO COMPARATIVE WORD-TABLE (contd) NAME MEANING ISM IRUM MA'ANA DUT IZEN NOMEN IKUR NOTIO MING LIBIZO II TALUSO NIMI NAMA ILMOITTAN ERTI NOM IGOA SENS UIGA NAME NOMBRE MEINUNG SENTIDO ONOMA JINA NOEMA MAANA SHEM PERU MASHMAO ARTHAMU NOV NAM ERTELEM MANA NOME DZINA SENSO MURERO NAMAE IGAMA IMI UMQONDO KNOW SAME 'IRIF ALDA SAWA KATUN JAKIN SCIO BER IDEM JENSHIH ZIBA HSIANGTUNG SWANA TIETAA KENAL SAMA SAMA SAVOIR ILOA MEME TUSA KENNEN CONOCER(SABER) SEIB MISMO GIGNOSCO JUA HOMOS ILE ILE YADA JNANA SHAVEH SAWA TUDOM BILMEK UGYAHAZ AYNI CONOSCERE UDIVHA STESSO FANA SHITTE AZI ONAZI NJALO MAN AT INSAN SARAM Fl E GIAZON HOMINEM DA AD JEN MUNNA CHIN FA IHMINEN MANUSIA SSA DI HOMME TAGATA A I MENSCH HOMBRE AN A ANTHROPOS MTU EPI PENYE ADAM MANUSHYUDA B VADDA EMBER ADAM A DA UOMO MUTHU A NGA OTOKO UMUNTU DE
In looking at the list, the question is what resemblances and differences can be observed between the different words used for each of the 43 meanings in the different languages. Before doing this, one must recall the warning given by Bloomfield (and many other philologists) against drawing any large conclusion about the relationships of languages from mere resemblance of vocabulary and in particular against attempting to construct new family relationships of language on the basis of scattered similarities of words. So Bloomfield, after discussing resemblances which can be explained by historical connections of languages, and accepting that there might be some resemblances resulting from universal factors in language, gave an illustration of the danger of basing too much on occasional similarities:
"Other resemblances between languages bear no significance whatever. Modern Greek MATI means 'eye' and so does the Malay word MATA. If we knew nothing of the history of these languages, we should have to work through their lexicons and grammars in search of other resemblances and then weigh the probabilities of historical connection, taking into account both the number of resemblances and their structural position. Actually our knowledge of the past forms bath of Greek and Malay shows us that the resemblance of the two words is accidental. Modern Greek MATI is a relatively recent development from an ancient Greek OMMATION 'little eye' and this word was in ancient Greek connected, as a secondary derivative, with an underlying word OMMA 'eye'. The Malay word on the other hand had in ancient times much the same phonetic shape as today. Even if against all present seeming, it should turn out some day that these two languages are related, the relationship would lie far back of Primitive IndoEuropean and Primitive Malayo-Polynesian times and the resemblance of the modern words for 'eye' would have nothing to do with this relationship".
With this caution in mind, one can now consider how far resemblances can be noted in the 23 languages between the words they use for the common ideas presented. The degree of resemblance can vary (in the same way as resemblance between words in any close language-grouping such as that of the Romance languages can vary). Resemblance can range from complete identity of sound, as in English HAND German HAND, through near identity, as in French COEUR Italian CUORE, a clear general resemblance, as in German ZUNGE English TONGUE, resemblance of structure, as in French LANGUE English TONGUE, resemblance of principal sound or sounds, as for LINGUA LISAN LIDAH (tongue), to remoter though possible resemblance, as for English HEART Greek KARDIA to complete dissimilarity, as for English NECK German HALS.
Given this varying degree of resemblance and the considerable quantity of material to be analysed, the way adopted here of tackling the description of the degree of similarity observed is to list the 43 words in the order which results from assessing the overall degree of similarity found between words drawn from the 23 languages i.e. the word where there is the greatest resemblance between the different languages is placed first and that where there is the least is placed last.
Proceeding in this way, in terms of the overall degree of resemblance, the words are placed in the following order:
CRAB LIP BODY
CUT MEANING MOUTH
LICK THROAT EYE
TONGUE HIT HIP/THIGH
NAME SAME NECK
WAVE SPIT PRICK
HEART EAR BREAST
AIR KNEE EARTH
DEAD WATER CHIN/JAW
RUB GO FACE
AT HAND CHEEK
NOSE VOICE SHOULDER
FIRE TOOTH GIVE
MAN HEAD
ELBOW KNOW
The words for CRAB can be grouped in order of their degree of
resemblance as follows:
CRAB SARATAN HSlEH
CRABE SARTAN
KRABBE RAPUELAIN
CAMBARO
CANCER TENGERIRAK
GRANCHIO
KARKINOS PA'A
KAPPU
KARAMARRI YENGEC
KANI
KAA DAMBATSHEKWA
KE
KEPITING
NKALA
INKALA
Out of 23 words for CRAB, 14 begin with K (or C) or have K as a prominent letter in the word (NKALA and INKALA); GRANCHIO is included for its general resemblance to other words in the group. Apart from these 15 words, 2 of the remaining 8 words resemble each other (SARATAN and SARTAN), both of which are Semitic (from Hebrew and Arabic). The remaining 6 words appear completely isolated (words from Chinese, Finnish, Hungarian, Samoan, Turkish and Venda). The 15 words which show a varying degree of resemblance to each other include those drawn from the 7 Indo-European languages (though the words do not appear to be related by descent or borrowing except for CRAB CRABE and KRABBE in English, French and German). The other IndoEuropean languages have chosen as words for CRAB ones which have some broad degree of resemblance but not identity e.g. Spanish CAMBARO and Italian GRANCHIO. It may be thought that it is coincidence that so many of the words for CRAB should begin with K (or C), a sound that happens in English to be associated with the set of words such as CLASP, CLAW, CRUSH &c. It appears more than a coincidence if one notes the rarity (as regards the other 42 words listed) of finding K (or C) as the initial letter or as a prominent letter. The only other word which shows a similar concentration of the letter K (or C) in the different languages is CUT. The broad resemblances between words drawn from unrelated languages are of interest:
Korean KE Japanese KANI
Swahili KAA Latin CANCER
Telegu KAPPU
Malayan KEPITING
The words NKALA and INKALA are drawn from two related languages, Lozi and Zulu.
The degree of resemblance shown seems well above chance and not to be explained as a matter of accident or coincidence, nor is it apparently due to borrowing - since one would expect borrowed words to resemble each other more closely than most of these do. The resemblance suggests that speakers of different languages in some way come to associate with the very distinctive shape and character of the CRAB sounds which tend to be the same or similar.
In some other languages (not systematically covered in the above examination) the words for CRAB seem to fit the same pattern e.g. Pahlavi KARZANG Portuguese CARANGUELO.
A similar examination can be made of words for CUT. The words can be arranged in terms of their resemblance as follows:
KO' SCHNEIDEN
KOYU
CUT VAG
CORTAR
COUPER TAGLIARE
KATA
KATA PEGI
KARAT
KEIRO
KIRU PUMA
SIKA
PIKATU MEMOTONG
LEIKATA
CAEDO TIPI
KESME
UTUMULA
Out of the 23 words for CUT, 15 begin with K (or C) or have K as a prominent letter in the word (SIKA PIKATU and LEIKATA). Apart from these 15 words, the remaining 8 words appear isolated from each other and from the main group (words from German, Italian, Hungarian, Korean, Lozi, Malay, Samoan, and Venda).
The 15 words which show a varying degree of resemblance to each other include words drawn from 5 Indo-European languages (though they all appear etymologically unrelated). French COUPER and English CUT are not normally related to each other etymologically (even though there is also COUTEAU in French for the instrument used for cutting - a knife).
Apart from the general resemblance of the words in the group (based on the letter K (or C) ) there are some closer resemblances which are of interest:
Telegu KOYU Arabic KATA
Chinese KO Swahili KATA
Japanese KIRU Basque PIKATU
Greek KEIRO Finnish LEIKATA
Hebrew KARAT
The degree of resemblance seems extraordinary and well beyond any explanation in terms of chance or coincidence. Some resemblance, for example between Arabic and Swahili, may be due to borrowing though it would be surprising if a word for an action as common and necessary as CUT should not usually be found as an aboriginal constituent of every language. The resemblances between Japanese, Greek and Hebrew cannot be explained in this way.
That the resemblance (and the predominance of words beginning with K or with K as a prominent letter) cannot be due to chance is confirmed by examination of words for CUT in some other languages. In Aranda (an Australian aboriginal language) the word for CUT is KAMA, Amerindian words are KAWATE SIKATE UKUTA KOODE SIIKI, Punjabi KET, Icelandic SKERA, Hawaian OKI, Egyptian NEKA.
The next word considered is LICK. The words can be arranged in order of resemblance as follows:
LICK NAKU MENJILAT
LIKEK
LIKHMADZO NAMERU SALOSALO
MILIKATU
LECKEN NANZWA KHOTHA
LECHER
LECCARE
NYAL
LIHIS
LINGO
LAZWA
LAMBA
LAMER
YALAMAK
Out of the 22 words listed, 14 begin with L or have L as a
prominent letter in the word. It is of interest that another 4
start with N (including NUOLLA which also has L as a prominent
sound). Out of the 14, 7 are drawn from Indo-European languages,
though several distinct roots appear to be involved. The
resemblance of the first 10 words is striking; besides English
LICK, these include:
Hebrew LIKEK
Basque MILIKATU
Hungarian NYAL
Arabic LIHIS
Greek LIKHMADZO
Latin LINGO
Other striking resemblances are:
Swahili LAMBA
Spanish LAMER
Turkish YALAMAK
Again the overall degree of resemblance seems much greater than can be explained as a matter of chance. There seems little likelihood of borrowing between the remote languages which show strong resemblances. As in the case of the words already considered for CUT and CRAB, a cursory survey of some other languages throws up similar forms. In Amerindian languages there is the word LAMBI for LICK (and for tongue LISIKI LAKSI LINGI), in Russian LIZAT, Albanian LEPIJ, Somali LEEF.
The next word considered is TONGUE. The words for tongue can be arranged in order of resemblance as follows:
LINGUA ZUNGE MIHI
LINGUA TONGUE
LENGUA SHE
LANGUE SITA
NALUKA
LISAN HYO
LASHON
GLOSSA
LIDAH
DIL
ULIMI
LILIMI
LULIMI
LAULAUFAIVA
NYELV
KIELI
Out of the 23 words, 14 have L as the initial letter or as a prominent letter in the word; another two also have L as a less significant letter. Two words appear completely isolated (from Basque and Korean); two resemble each other in form (German ZUNGE and English TONGUE - etymologically related) and also resemble in form a number of the words beginning with L on the pattern of LINGUA LANGUE. The remaining two words which may have some resemblance are drawn from Japanese and Chinese.
The large group of words beginning with L contains words from many language families. Besides the 5 Indo-European words, NALUKA is Telegu, LISAN and LASHON are Semitic, LIDAH and DIL are Malay and Turkish and the remaining three words are from the Bantu languages.
The degree of resemblance seems remarkable. The concentration of words beginning with L is matched only by the group of words for LICK (which seem closely related to words for TONGUE) and to a lesser extent by words for LIP. Apart from these, L is rare as an initial letter for the 43 words under consideration. The concentration seems to go beyond anything that could be the result of chance or coincidence. Words from other languages not included in the Table show a similar concentration in the use of L : Amerindian words for tongue have already been quoted LINGI NILI LAKSI LISIKI. In addition there are the Hawaian ALELO and the Aranda LINGA.
The next word considered is NAME. The words from the different languages can be arranged in order of resemblance as follows:
NAME ISM IGOA PERU
NAME IZEN IGAMA
NAMA SHEM
NAMAE
NAM DZINA
NOM JINA
NOME
NOMEN IRUM
ONOMA
NOMBRE LIBIZO
NOV
NIMI
MING
Out of the 23 words for NAME, 12 begin with N and a further 8 have N or M as a prominent letter (including MING which has both N and M though in reverse order). One word PERU appears completely isolated and two others IRUM and LIBIZO are probably isolated (at any rate in terms of the languages dealt with here). These three words are from Telegu Korean and Lozi respectively.
The 12 words in the main sub-group beginning with N include, besides those from the 7 Indo-European languages, words from unrelated language-families, Finnish, Hungarian, Japanese, Malay and Turkish. Some of the resemblances between unrelated languages are striking:
NAME English (and German)
NAMAE Japanese
NAM Turkish
NAMA Malay
NIMI Finnish
ISM Arabic
IZEN Basque
SHEM Hebrew
IGOA Samoan
IGAMA Zulu
DZINA and JINA are from two related languages, Swahili and Venda.
The concentration of words beginning with N or including N as a prominent letter is unusual and remarkable. For the other words in the list, N is rare as an initial letter except for words for NOSE. It is unlikely that the close resemblances between remote languages can be due to borrowing; not only is this on the face of it improbable (Japanese NAMAE can hardly be a borrowing from English or German) but the idea NAME is something one would expect to be found in every language as a basic aboriginal element. Nor is it likely that Malay NAMA is borrowed either from the Japanese or the English.
A rapid examination of a number of other languages shows that the result described above is not dependent on pre-selection of the languages used. In Amerindian languages one finds MAMI WAMI NIK, Vietnamese MING, Russian IMYA, Hawaian INOA, Albanian EMER, Norwegian NAVN, Punjabi NAN, Icelandic NAIN, Pahlavi NAM.
Rather than examine at the same length each of the remaining 38 words, it is proposed simply to draw attention to some of the more striking resemblances which, cumulatively with the other material presented, seem to make it impossible to explain the overall degree of resemblance between the different languages in their words for common objects and actions a matter of sporadic resemblance or chance.
HEART Basque BIHOTZ
German HERZ
English HEART
Telegu HRDAYAMU
Japanese KOKORO
Latin COR
Italian CUORE
to which can be added Egyptian HAT.
AIR
Arabic ARYAH
Greek AER
Hebrew AVIYR
Samoan EA
Swahili HEWA
DEAD Arabic MAIYIT
French MORT
Hebrew MET
Lozi MUSHU
Malay MATI
Samoan MATE
Swahili MUFU
Venda MUFU
to which can be added: Russian MYORTVYI Egyptian METU Hawaian
MAKE Punjabi MERYA.
NOSE English NOSE
German NASE
French NEZ
Finnish NENA
Japanese HANA
Venda NINGO
plus Amerindian NASA OSA NISI Russian NOS Icelandic NEF
Norwegian NESE.
MAN English MAN
Lozi MUNNA
Malay MANUSIA
Telegu MANUSHYUDA
Venda MUTHU
Zulu UMUNTU
Swahili MTU
MEANING English MEANING
Arabic MA'ANA
Swahili MAANA
Turkish MANA
German MEINUNG
HIT English HIT
Hebrew HIKAH
Venda HIRWA
Hungarian UT
Japanese UTU
Samoan TA
Lozi NATA
Zulu BETA
plus Amerindian ETI UKA
Spanish BEGAR
Samoan PIGA
plus Amerindian PEKA
SAME English SAME
Finnish SAMA
Malay SAMA
Lozi SWANA
Arabic SAWA
Telegu SAWA
Hebrew SHAVEH
plus Russian SAMYI Icelandic SAMUR
EAR English EAR
German OHR
Chinese ERH
French OREILLE
Greek OTA
Hebrew OZEN
It is proposed to deal more fully with the words for EYE which
are of special interest. The words can be grouped in terms of
resemblance as follows:
EYE OCCHIO ME NUN
YEN OCULUS OMMA
AYIN OJO MATA LIITO
'AIN AUGE MATA
YEUX BEGI MATO KANNU
GOZ
ISO SILMA
JICHO
Three of the 23 words appear completely isolated (from Korean, Lozi, Telegu). The remaining words are placed in three sub-groups within which the degree of resemblance varies. The first sub-group brings together words from some remote languages: English EYE Chinese YEN Hebrew and Arabic AYIN and 'AIN and French YEUX (the plural form of OEIL which in form and etymological origin is linked to the words in the second sub-group beginning OCCHIO but in actual pronunciation is closer to the first sub-group beginning EYE).
The second sub-group brings together the languages of Latin origin (Italian, and Spanish but not French) with the German AUGE and, as more remote structural similarities, Basque BEGI and Swahili JICHO. See also Amerindian OGI with German AUGE and Latin OCULUS.
The third group is of particular interest. The Japanese ME is followed by four very similar words:
OMMA Greek (Modern Greek MATI)
MATA Malay
MATA Samoan
MATO Venda
In addition to these, from other languages there are Vietnamese MAT (Egyptian MAA to see) Hawaian MAKA.
Bloomfield's comment quoted earlier that the resemblance between Modern Greek MATI and Malay MATA is accidental and entirely without significance seems somewhat less justified when the words are placed in a larger group including remote languages such as Venda and Vietnamese.
Of the other comparisons, that between English EYE and Chinese YEN may seem extraordinary; the resemblance between English EAR and Chinese ERH makes it even more extraordinary. At some point explanation by way of coincidence begins to seem incredible - but no explanation in terms of language relation seems possible, any more than for the resemblance between Venda, Malay and Greek.
WATER The words can be grouped as follows:
WATER MADI UR SU
WASSER MAJI AYER SHUI
VESI MEZI AQUA
VIZ MIZ ACQUA HUDOR
VAI MWAI AGUA
MAYIM NILLU
AMANZI EAU
MUL
In terms of resemblance the three sub-groups seem well-defined, though there are some resemblances between words in the different sub-groups e.g. VAI and MWAI1 VIZ and M1Z, VESI and MEZI, and a general resemblance of structure between the sub-groups.
There are two words which appear isolated (from Greek and Telegu). SU and SHUI are Turkish and Chinese respectively. The sub-groups bring together words resembling each other from remote languages: VAI is Samoan, VIZ and VESI Finno-Ugrian with some resemblance with German WASSER and English WATER.
MADI MAJI and MEZI are from the Bantu languages but MIZ is Japanese and MWAI and MAYIM are Semitic. UR is Basque, AYER Malay, AQUA Latin. The French EAU is related etymologically to AQUA but in sound, as currently pronounced, is quite distinct
Other languages show resemblance to one or other of the sub-groups - sometimes to two sub-groups at the same time as is the case for the Australian Aranda word KWATJA. This has a resemblance to the group beginning WATER, and to the group including ACQUA. Amerindian words include AKWINI as well as MA and UMA. There are also Swedish VATEN Hawaian WAI Russian VODA Norwegian VANN; and Egyptian MU (Korean MUL Arabic MWAI).
HEAD
The words for HEAD can be grouped as follows:
CAPO RAS
KOPF ROSH
CAPUT
KEPALA BASH
KEPHALE
CABEZA
KICHWA
Of the 20 words listed, other than the 10 above, the remainder appear to be isolated including French TETE and English HEAD. There seem in most languages to be a number of synonyms for HEAD; listing of the synonyms would bring out a greater degree of resemblance not only between remote languages but between languages belonging to the same family, particularly the Indo-European.
Nevertheless, the words in the first sub-group come from a wide spread of languages:
Italian CAPO, German KOPF, Malay KEPALA, Greek KEPHALE and Spanish CABEZA. KICHWA is Swahili.
The resemblance between the Malay and some Indo-European words is striking.
Survey of other languages produces some interesting resemblances: KAPUTA Australian aboriginal CAPUT Latin KUPA Amerindian; HEWIDA Amerindian HEAD English HOOFD Dutch; Egyptian TATA French TETE. There is also Amerindian KAPPALI (forehead) which closely resembles Greek KEPHALE and Malay KEPALA.
BODY
Basque GORPUTZ English BODY
Latin CORPUS Arabic BADAN
German KORPER
The Basque word may be a borrowing, though it is surprising
if
Basque had no word of its own. The resemblance between English
and Arabic is of interest.
NECK
The words can be grouped as follows:
NECK COU MULAIA
NYAK KUBI UMQALA
'UNK KAULA
COLLO
AUKHEN COLLUM
CUELLO
MOK
TENGKOK
The first two sub-groups include 12 out of the 23 words. The remaining words (other than the two Bantu words (MULALA and UMQALA) appear to be isolated (including the German HALS, though etymologically this may be related to COLLUM).
In the first sub-group, the resemblance of the Arabic 'UNK, Hungarian NYAK and English NECK is striking. There seems no possibility of borrowing or explanation in terms of language-relation. The Greek AUKHEN is included because the main sounds N and K are the same though in reverse order; metathesis, the changing of the order of sounds in words, is a common phenomenon.
In the group beginning COU, most of the words are etymologically related to the Latin COLLUM but KUBI is Japanese and KAULA is Finnish. Though borrowing of the Finnish word from an Indo-European language has been suggested (from one of the Baltic languages), this seems unlikely. In Finnish, as in other languages, there is a parallelism between words for functionally similar parts of the body. Besides KAULA (neck) there is NILKKA (ankle). Similarly in English the NK grouping appears in ANKLE KNEE NECK KNUCKLE (and possibly a modified form in FINGER). In Japanese, there are HIZA and HIZI meaning knee and elbow respectively, in Arabic KU and KAB meaning elbow and ankle, in French COU and COUDE meaning neck and elbow, in Hungarian NYAK and KONYOK meaning neck and elbow, in Japanese KUBI and ASIKUBI meaning neck and ankle, in Samoan TAPUVAE TULIVAE TULILIMA meaning ankle knee and elbow respectively, in Zulu IQAKALA UMQALA meaning ankle and neck.
The resemblances amongst the words for NECK seem to go well beyond what can be explained as a result of coincidence, and with the comparisons between words for other joints (knee, ankle, elbow), there is a clear impression of some well-organised underlying system of relation between the sounds used and the meanings. A survey of words used in some other languages supports this: fitting the group of words beginning with COU there is Vietnamese CO and for the group of words beginning NECK, there are Amerindian NUKI NUK TENUK NOXI plus words in Dutch and German for nape of the neck NEK NACKEN) and Egyptian NEHEBET.
EARTH
Some resemblances are:
EARTH English ERDE German
ARD Arabic ODAMAH Hebrew
However, in most languages there are several synonyms or near-synonyms for EARTH, sometimes with slightly different applications (as in English EARTH GROUND LAND SOIL WORLD) Examination of the range of synonyms in other languages would be likely to produce a larger number of resemblances. To some extent, the words in different languages for EARTH AIR WATER and FIRE show interrelations similar to those seen between words for different joints of the body considered immediately above. So there are: MUL PUL in Korean meaning water and fire, UR LUR in Basque meaning water and earth, ARD ARYAH in Arabic meaning earth and air, MAA ILMA in Finnish meaning earth and air, VIZ TUZ in Hungarian for water and fire, NIPPU NILLU in Telegu for fire and water.
FACE
There is the resemblance:
MUKAU Japanese
MUKA Malay
MUKHAMU Telegu
SHOULDER
Compare:
KATEF Hebrew and OMUZ Turkish
KATA Japanese OMOS Greek
KITF Arabic (HUMERUS) (LATIN)
GIVE
There are the following resemblances:
AD FA
'ADA PA
DAR
DARE
DONNER
DIDOMI
ANTAA
NATAN
In the first sub-group, AD and 'ADA are Hungarian and Arabic, DAR is Spanish and ANTAA and NATAN are Finnish and Hebrew. FA and PA are both from African languages.
From other languages, there are: NDAMA Australian aboriginal IDAI IDE from Amerindian languages TA Egyptian DADAN Pahlavi DE Punjabi DAVAT Russian.
The conclusions drawn from the above examination of resemblances between words in 23 languages for 43 common objects, actions etc. are the following:
(a) In the case of many, if not most, of the common words examined, there are resemblances between words drawn from remote languages which seem to go well beyond chance;
(b) these resemblances cannot plausibly be explained as a result of borrowing between languages or of previously unobserved relationships of descent between broad language families;
(c) there is a clear probability that, for well-defined referents, there is a tendency for a limited range of sounds to be used to express them - which results in a greater than chance resemblance between languages;
(d) in the case of some words, there are striking and almost complete resemblances between very remote languages; in other cases there is a tendency for a particular letter to be used as the initial letter or as a prominent letter in most of the languages for the particular referents;
(e) there are also signs of internal relationships between the sounds used in any given language to compose words referring to referents which are themselves related in function or form;
(f) these findings are consistent with and positively support the hypothesis that there is a systematic underlying sound-meaning relation in language and that this relation is manifested in many if not all languages The evidence is consistent with the experimental study of sound symbolism described in the previous chapter which demonstrated the ability of experimental subjects to reach conclusions with a greater than chance accuracy about the meanings of words in languages unknown to them, in the conditions of the experiments reported;
(g) the findings are not in conflict with the views of comparative linguistics on the family relationships of language which are based in a very similar way on the observation of resemblances in structure and vocabulary which are thought cumulatively to be so great that no other explanation (borrowing, convergence, analogy, accident and coincidence) is plausible as an explanation. Family relationships between languages (i.e. variations in form which have developed -in the same way as dialects develop - over time by different sections of an originally united language-community) are fully compatible as one form of explanation (using systematic sound-laws in parallel to systematic dialectal variations within a single language) with the assumption of a more generally shared natural underlying relationship between sound and meaning which increases the probability between languages that words for similar referents will be similar in sound-composition;
(h) an explanation on these lines supplements accepted comparative linguistics and provides an answer to some embarrassing peculiarities which have had up to now to be dismissed as the result of chance. So, because they cannot be fitted into the system of the Indo-European sound-laws, the following resemblances:
HAVE English DAY English PATH English
HABEN German TAG German PATH Persian
HABEO Latin DIA Spanish
AVERE Italian DIES Latin BAD English
BAD Persian
WHOLE English FIRE English
HOLOS Greek FEUER German
FEU French
FOCUS Latin
PUR Greek
have had to be dismissed as meaningless accidents. Once it is recognised that there can be resemblances between languages which derive from a natural underlying basis and not necessarily from common language-descent, it is not surprising to find words such as the above which resemble each other between genetically related languages but do not follow the rules of dialectal variation which otherwise link a large part of the vocabularies.
The system of comparative linguistics was built up by not dismissing as accidental observed resemblances between languages and by the cumulative evidence that in this way was gathered in support of the hypothesis that there were underlying regularities between related languages. The system of verification of any relation between two languages involved "a hierarchy of criteria through which the comparative linguist will descend to find more and more support (or possible refutation) for a given hypothesis... To escape from all possible ambiguity, comparative linguistics has to base any proof of common origin and inference of proto-language forms on words that cannot conceivably be the product of coincidence brought about by extra-linguistic factors... The hierarchy will proceed through the more 'colourless' forms such as pronouns, articles, flexional endings, until these have been exhausted.... (the search must continue for regular resemblances) until "It is fantastically improbable that coincidence could have created these regular correspondences". (Lord - Comparative Linguistics).
Applying what is said about comparative linguistics to the results set out in this section, it is fantastically improbable that coincidence could have created the degree of resemblance observed between unrelated languages. The words used in the examination have been ones that are unlikely to be borrowed between languages. The accumulation of evidence makes it increasingly likely that some other principle, besides that of genetic relation between languages, must be introduced to explain the uniformities observed. The only gap so far is that the examination of resemblances has not extended to 'colourless' forms such as pronouns, articles, demonstratives or to words for basic colours on which much of the structure of comparative linguistics has been built. These other areas are dealt with in the remainder of this Chapter.
The words dealt with in this section are the demonstratives THIS and THAT and the personal pronouns I (ME) THOU YOU HE IT WE THEY. The pronouns are also demonstratives in the sense that in any particular circumstances the reference of the pronouns must be indicated. So 'I' means 'THIS PERSON HERE who is speaking', 'YOU' means 'THIS PERSON HERE to whom I am speaking', 'HE' means 'THAT PERSON THERE about whom I am speaking'. In one sense demonstratives and pronouns have the most sharply defined meanings of any words - they refer to precisely identified persons or objects - but in another sense the demonstratives have no meaning other than is involved in the act of demonstration, the act of pointing to somebody or something; in this respect they resemble purely grammatical forms like the prepositions and conjunctions. In contrast, other words, nouns like CHAIR or WOMAN, are much more specific than demonstratives since they refer only to a particular idea; these other words can be used meaningfully in isolation from a context to convey the idea of a particular material or living structure.
Because demonstratives and pronouns are in the sense indicated free of specific content, they have the most uniform application throughout the world. They are a basic part of all languages because no language can be conceived of which does not make possible a distinction between the speaker and the person addressed and which does not allow indication of the particular real objects or persons referred to, This is why demonstratives and pronouns are of interest for study of the relationships between languages. Study of pronouns played an important part in the schemes of family relationships between languages constructed in the nineteenth century since it was assumed that for pronouns (and indeed for demonstratives more generally) it was more unlikely than for other words that borrowing rather than descent could explain observed resemblances. It was thought that pronouns and demonstratives must be primitive elements in any language which could be expected to descend from the parent to successor languages (as in the Indo-European family). Jespersen commented: "Personal and demonstrative pronouns, articles and the like, are scarcely ever taken over from one language to another, They are so definitely woven into the innermost texture of a language that no one would think of giving them up."
This is not to say that all languages necessarily have the same number and classification of pronouns and demonstratives. Each language can be expected to distinguish I and YOU but some languages do not distinguish between IT and THAT or between HE and THAT MAN. Nor do all languages distinguish YOU singular (THOU) and YOU plural or even I singular and WE plural - (in some languages WE may be used in place of I, YOU may be used for singular and plural). Some languages have additional distinctions of pronouns and demonstratives which are not found in English. They distinguish WE (including the person spoken to) from another WE (not including the person spoken to). In some African languages, Homburger points out that between two people speaking to each other the role of the pronouns depends on whether it is a matter of speaking of a person or object which is present or absent, known or unknown, definite or indefinite. In a number of Oriental languages, a bewildering variety of forms is used for first and second personal pronouns depending on the relative status of the speaker and the person addressed, with elaborate circumlocutions for politeness. There may, even in European languages, be fine distinctions such as in the French use of TU and VOUS or the German DU and SIE. Similarly the third personal pronoun or some form substituting for it may be used in politeness in place of the first or second personal pronouns as in the roundabout phrases used in Spanish and Italian USTED LEI.
For the demonstratives as such, some languages distinguish more shades of meaning than in English. Rather than two words simply equivalent to THIS and THAT, they have additional words expressing more precisely the distance or relative position of the person or object referred to. So there may be words for 'THIS near me' 'THIS near you 'THAT near at hand' 'THAT far away or absent'. French has something on these lines with the distinction in use between CE CELUI-CI CELUI-LA.
In many inflected languages, there appears to be some systematic relation between the pronouns in isolation and the forms of the inflection used to convey them (the suggestion that inflections could have developed from so to say impacted pronouns is one which has regularly been advanced over many years). There may be some relation between demonstratives and pronouns and the forms of the verb TO BE; the verb TO BE itself has a certain demonstrative character. It may not be fanciful to see some significant resemblance between the demonstratives (words for THIS and THAT) and the 3rd person singular of TO BE in a number of languages. ESTI (Greek) EST (Latin) ESTA (Portuguese) meaning IS have some resemblance to TIS ISTE ESTE (demonstratives in these languages).
The conclusion from this discussion is that although in principle the essential meanings to be conveyed by demonstratives and pronouns are of the greatest uniformity and clarity throughout the world, the overall picture of the forms and use of demonstratives and pronouns is not necessarily simple. The discussion was undertaken as a preliminary to examination of the forms which demonstratives and pronouns are found to take in a wide range of languages.
In the remaining part of this section, the procedure in examining demonstratives and pronouns is similar to that followed earlier in the examination of resemblances between words drawn from a range of languages i.e. the collection of words from a number of languages is first presented, the apparent uniformities and resemblances between words from the different languages are noted and the extent to which the systematic resemblances observed are compatible with the theory of the direct relation of sound and meaning presented in this book is considered.
The material contained in the Table immediately following consists of words for pronouns and demonstratives drawn from some 70 languages; words used for the 3rd person singular of the verb TO BE are also included in view of the possible link to the demonstratives already referred to. Words are also included for the definite article THE.
For some languages, words are missing for some demonstratives or pronouns. This may be because the particular language does not make the distinction existing in English e.g. has no word for THE or IS or does not separate the word for HE from the word for THAT. As far as possible, the words presented for the pronouns are those which convey the basic simple meaning and honorific forms, forms of politeness, have been excluded.
The languages included in the list are from many different families. The languages are chosen to give the widest possible spread and include besides the more important modern languages, a number of important ancient languages such as Anglo-Saxon, Egyptian, Hittite, Latin, Pahlavi, Sanskrit, Sumerian. Given the range of languages and the room for dispute about transliteration from non-Roman scripts, there are bound to be some problems of orthography but it is thought that these do not vitiate the broad comparisons made.
PRONOUNS AND DEMONSTRATIVES
I THOU YOU HE IT
Albanian UNE TI JU AY ATE
Amerindian MI NI MA HA -
Anglo-Saxon IC THU GE HE HIT
Arabic ANA INTE INTU HU -
Aranda JINGA UNTA RANKARA ERA ERA
Arawak NUKA TIWA ITI LI IRORO
Basque NIK HI ZUEK - -
Bengali AMI TUMI TOMRA TINI IHA
Bulgarian AZ TI VI TOE TO
Burmese GNAH THIN NINDO THU THU
Chibchan NAN MA
Chinese WO NI NIMEN T'A T'A
Czech JA TY VY ON ONO
Danish JEG DU I HAN DET
Dutch IK JIJ U HIJ HET
Egyptian NUK TU TEN SU SET
English I THOU YOU HE IT
Eskimo VINGA ITLPIT ITLPITSE ITLE ITLE
Etruscan Ml -
Finnish MINA SINA TE HAN SE
French JE TU VOUS IL LE
Georgian ME SEN TKVEN ES -
German ICH DU SIE ER ES
Greek EGO SU HUMEIS AUTOS AUTO
Guahiban HAN HAM PAXAMI PONI -
Hausa INA KAI KU YI YA
Hawaiian AU OE OUKOU IA IA
Hebrew ANI ATA ATEN HOU OTO
Hindi MAI TU TUM YAH YIH
Hittite UK ZI SUMES
Hungarian EN TE ONOK 0 -
Icelandic EG THU THITH HANN THATH
Irish ME TU SITH SE -
Italian 10 TU VOl EGLI LO
Japanese WATASHI ANTA ANATA KARE SORE
PRONOUNS AND DEMONSTRATIVES
WE THEY THIS THAT THE IS
Albanian NE ATA KETE ATE AY
ESHTE
Amerindian ISA - HEEKE IMA - -
Anglo-Saxon WE HIE THES THAET SE IS
Arabic NIHNA HINNI HADA HADAK - -
Aranda NUNA ETNA NANA TANA ERA -
Arawak WIA HEI TIYAAHI ILE WA -
Basque GU - HAU HORI A IZAN
Bengali AMRA TAHARA E Al JE -
Bulgarian NI TE TOZI ONZI A E
Burmese GNADO THUIDO DI IDA - SHI
Chibchan - - - - - -
Chinese WOMEN T'AMEN CHE NA - SHIH
Czech MY ONI TENTO TEN - JE
Danish VI DE DENNE DEN EN ER
Dutch WIJ ZIJ DEZE DIE DE IS
Egyptian N SEN THEN PF PA AU
English WE THEY THIS THAT THE IS
Eskimo VANGKUTA ITLAIT UNA TAUNA - -
Etruscan - - - - - -
Finnish ME HE TAMA TUO - ON
French NOUS ILS CECI CELA LE EST
Georgian CVEN ESENI ES IS - -
German WIR SIE DIESER JENER DER 1ST
Greek HEMEIS AUTOI HOUTOS EKEINOS HO/TO ESTI
Guahiban WAXAI POMONE XUA BAHARA - -
Hausa MU SU WANNAN NAN N CE
Hawaiian MAKOU LAKOU KEIA KELA KA -
Hebrew ANOU HEME ZE HENE - -
Hindi HAM VE YAH JOVAH - HAl
Hittite WES - KAS ANIS - -
Hungarian MI OK EZ AZ - -
Icelandic VITH THEIR THESSI THETTA HINN ER
Irish SINN SIAD SO SAN AN IS
Italian NOI ESSI QUESTO QUELLO IL E
Japanese WATASHI-TACHI KARERA KONO SORE -DES
PRONOUNS AND DEMONSTRATIVES
I THOU YOU HE IT
Javanese AKU KAU KAMU IA -
Khoin AM A ANXA E HA
Korean NA TANGSIN - KIRI KUGOT
Latin EGO TU VOS ILLE ID
Lithuanian AS TU JUS - -
Malay AKU ENGKAU KAMU IA DIA
Malayalam NAN NI NI AVAN ATU
Manchu BVMINI SI SUVE I -
Mayan AYINE -
Mongolian BI TA TA TERE -
Norwegian JEG DU I HAN DET
Nubian Al ER IR TER TER
Pahlavi AZ TO ASMA AN -
Persian MAN TO SHOMA U -
Polish JA TY WY ON TO
Portuguese EU TU VOS ELE ISTO
Rumanian EU TU VOl EL -
Russian YA TY VY ON ETO
Samoan ITA E TOU IA NA
Sanskrit AHAM TVAM YUYAM - IDAM
Spanish YO TU VOSOTROS EL LO
Sumerian MA ZA MENZE ENE -
Swahili MIMI WEWE NINYI YEYE -
Swedish JAG DU NI HAN DET
Tagalog AKO KAYO IKAW SIYA ITO
Tamil NAAN NII NINGAL AVAN ATHU
Telegu NENU NIVU MIRU ATADU ADI
Thai CHUN - TAHN KOW MUN
Tibetan NA KYOD KYE KO KON
Tucanoan YII BII BI - -
Turkish BEN SEN SIE 0 ONU
Vietnamese TOI ONG ONG NO NO
Welsh Ml TI CHWI EF -
Yoruba EMI IWO ENYIN ON -
Zulu MINA WENA NINA YENA YENA
PRONOUNS AND DEMONSTRATIVES
WE THEY THIS THAT THE IS
Javanese KAMI MEREKA INI ITU --
Khoin BA ABE AA A - ESO
Korean URI KUDUL I KU - I
Latin NOS ILLI HIC ISTE - EST
Lithuanian MES - SIS ANAS - -
Malay KAMI MEREKA INI ITU - ADA
Malayalam NAM AVAR I A - UNTU
Manchu BE CE - - - BI
Mayan - WILl LA - -
Mongolian BITA TEDE ENE TERE - BAINA
Norwegian VI DE DENNE DEN EN ER
Nubian AR TIR IN MAN - -
Pahlavi AMA AWESAN ED AN - AST
Persian MA ISHAN IN KEH - AST
Polish MY ONI DEM TANTEN - -
Portuguese NOS ELES ESTE ESSE 0 ESTA
Rumanian NOI El ACEST ACELA EL -
Russian MY ONE ETOT TOT - -
Samoan MA LA LENEI LELA LE UA
Sanskrit VAYAM - AYAM ASAU - ASTI
Spanish NOSOTROS ELLOS ESTE ESE EL ES
Sumerian MENDE ENENE E NE BI -
Swahili SISI WAO HIKI LE - NI
Swedish VI DE DENNA DENDAR DEN AR
Tagalog KAMI SILA ITO YAON ANG AY
Tamil NAM AVANGAL IDI ADI - IRUKU
Telegu MEMU VARU VIDU VADU - UNNADU
Thai ROW KOW NII NAN - CHY
Tibetan KOU - DI DE CIG -
Tucanoan BARI IDA ATE SISE - -
Turkish BIZ ONLAR BU SU - DIR
Vietnamese CHUNG TOI CHUNG NO CAINAY CAIKIA - LA
Welsh NI WY HWN HWNNW Y YW
Yoruba AWA AWOU EYINI IYEN NA WA
Zulu THINA BONA LELI LOKHO - -
I
The words for the first personal pronoun 'I' in the different languages can be grouped in terms of resemblance as follows (where a word resembles more than one form, it is repeated with the other words which it resembles):
I IC EG ME EN BI AS ITA
Al 1K EGO ME NA BI AZ
AU UK EGO ME NA BEN AZ TOI
EU ICH JEG Ml NAN
EU AKU JEG Ml NAN HAN
IO AKU JAG Ml NAAN
YO NIK JINGA EMI NENU WATASHI
YA NUK MIMI ANA
YII NUKA VINGA MINI ANI WO
JA MINA AYINE
JA MINA INA
JE MA UNE GNAH
JEG MAI
JEG MAN
JAG AM
AHAM
JINGA
63 of the words given in the Table for 'I' can be grouped with other words which they resemble more or less closely. Only 7 words appear completely isolated from the rest; these words ITA TOI HAN WATASHI WO CHUN GNAH mean 'I' in Samoan, Vietnamese, Guahiban, Japanese, Chinese, Thai and Burmese respectively. The remaining words are formed into 5 large groups on the basis of resemblance and 2 small groups (each containing only three words). Between some of the large groups, the degree of dissimilarity is not very large; so words in the group beginning EG could easily be linked with words in the group beginning IC. Similarly the differences between the words in the group beginning EN and the words in the group beginning ME are not large Words in the group beginning I could mostly be linked with words in the group beginning IC or in the group beginning EG Indeed it could be said that, apart from the six words in the two small groups and the five isolated words, all the remaining 59 words could be arranged in only two large groups, one containing the groups beginning with I IC and EG and the other containing the groups beginning ME and EN (it might even be possible to associate the small group BI BI BEN with this second large group). This is an extraordinary result given the range of languages considered and strongly suggests some underlying principles of uniformity in the formation of the sounds appropriate for naming the first personal pronoun 'I'.
Examination of the groups of words set out above in more detail throws up some interesting resemblances across language-families. In the first main group, that beginning with I there are the following:
I English JEG Danish
Al Nubian JAG Swedish
JINGA Aranda
AU Hawaiian
EU Portuguese
EU Rumanian
JE French
YII Tucanoan
I0 Italian
YO Spanish
In the second main group, there are:
IK Dutch
IC Icelandic
UK Hittite
AKU Javanese
NIK Basque
NUK Egyptian
NUKA Arawak
In the third main group, there are:
JINGA Aranda VINGA Eskimo
In the fourth main group, there are:
ME Irish MA Sumerian
ME Georgian MAI Hindi
Ml Etruscan MAN Persian
MI Other Amerindian AM Khoin
EMI Yoruba AHAM Sanskrit
MIMI Swahili
MINI Manchu
MINA Finnish
MINA Zulu
This is a really remarkable collection of resemblances. It suggests that there is something especially natural and appropriate about the use of the sound M for referring to the speaker himself. In the fifth group, there are:
EN Hungarian INA Hausa
NA Korean UNE Albanian
NA Tibetan
NAN Malayalam
NAN Chibcham
ANA Arabic
ANI Hebrew
AYINE Mayan
The words in the small groups are:
BI Manchu AS Lithuanian
BI Mongolian AZ Pahlavi
BEN Turkish AZ Bulgarian
YOU (THOU) The words for the second personal pronoun (singular) in the different languages can be grouped in terms of resemblance as follows:
TI DU NI KAI HI E TANGSIN TI DU NI KAU A TI DU NI ENGKAU ER IWO TY DU NII HAM OE TY SI NIVU KYOD ONG TY SU TU ZI BII TU ZA TU SEN ITLPIT TU SEN TU SINA JIJ TU TU THIN WEWE TU TU MA TIWA THU WENA THU THOU TA TA TE TO TO ATA ANTA UNTA INTE TUMI TVAM
57 of the words given in the Table can be grouped with other words which they resemble more or less closely. Only 9 words appear completely isolated; these words TANGSIN IWO ONG BII ITLPIT JIJ WIEWE MA WENA mean THOU in Korean, Yoruba, Vietnamese, Tucanoan, Eskimo, Dutch, Swahili and Zulu respectively. The remaining 62 words are formed into one extremely large group, one medium-sized group and four small groups each of which contains 3-5 words. The two largest groups resemble each other quite closely; these groups beginning TI and DU could plausibly be formed into one large group containing 40 words. Again as for words meaning 'I', this is an extraordinary concentration of resemblance and strengthens the case for believing in uniform principles operating across languages to determine the most appropriate sounds to express the idea 'YOU (THOU)', the 2nd person singular.
It is apparent even without giving examples that the same sounds must be used for THOU in remote languages. The following are of particular interest:
TI Albanian DU German NI Chinese
TI Welsh SU Greek NI Malayalam
TU French SI Manchu NI Other Amerindian
TU Lithuanian ZI Hittite
TIWA Arawak ZA Sumerian
TE Hungarian KAI Hausa
TY Russian SEN Georgian KAU Javanese
TA Egyptian SEN Turkish
TA Mongolian SINA Finnish
ATA Hebrew
ANTA Japanese
UNTA Aranda
TO Persian
Of the first examples, except for the 5 words from Indo-European languages, the remainder come from completely unrelated and geographically remote languages. The resemblance NI between Chinese, Malayalam (one of the Dravidian languages of Southern India) and the Amerindian word is also striking.
YOU (plural) The words can be grouped as follows:
I TE VI NI KU GE
I TA VY NI KYE
U TOU VY NINYI OUKOU RANGKARA
YOU ITI VOl NIMEN IKAW
YUYAM INTU VOl ENYIN ITLPITSE
JU TEN VOUS NINGAL KAMU
JUS TUM VOS KAMU ONOK
ZUEK TAHN VOS NINDO
SIE ATEN VOSOTROS IR
SITH TOMRA WY
SUVE ANATA CHWI ANXA
SUMES THITH
SIZ BI TKVEN
SHOMA
ASMA MENZE
HUMEIS MIRU
ONG
PAXAMI
MA
A rather larger number of words for YOU (plural) seem to be isolated, 12 in all. The remainder of the words can be formed into 3 large groups and two smaller on the basis of resemblance. Each of the groups seems distinct with no observable overlapping with other groups. There is some resemblance to words already listed for YOU (singular); the group beginning TE clearly resembles words in the large group beginning TI of words for YOU (singular). There is also some resemblance between words beginning with an S or Z and words beginning with S and Z for YOU (singular). For both YOU (singular) and YOU (plural) there are groups beginning with NI. Some of the words are simply plural forms Similarly there is a resemblance between the group beginning KU YOU plural) and that beginning KAI YOU singular). However, the group beginning VI seems specific to the plural YOU with no words resembling it found for YOU singular (other than the Tucanoan word BI which goes with the similar Tucanoan word BII YOU singular).
Interesting examples of resemblances are the following:
JU Albanian TE Finnish NI Malayalam
JUS Lithuanian TEN Egyptian NI Swedish
ZUEK Basque ATEN Hebrew NIMEN Chinese
SIE German TUM Hindi NINYI Swahili
SUMES Hittite TOU Samoan
SHOMA Persian TA Mongolian KU Hausa
HUMEIS Greek TAHN Thai KYE Tibetan
ITI Arawak OUKOU Hawaian
SUVE Manchu IKAW Tagalog
WE The words for the 1st person plural in the different languages can be arranged as follows in terms of resemblance:
WE ME N KAMI HAM
WE MES NE KAMI
WES Ml NI KAMI GNADO
WIJ MU NI
AWA MY NOI KOU CVEN
WIA MY NOI GU
WIR MY NOS VANGKUTA
VI MA NOS SISI
VI MA NOUS ISA URI
VI AMA ANOU SINN
VITH AMRA NUNA BARI
VAYAM MAKOU NIHNA
MEMU NOSOTROS ROW
BE NAM
BA HEMEIS NAM THINA
BIZ
BITA MENDE CHUNG TOI
WOMEN WATASHI-TACHI
WAXAI
10 of the words for WE cannot in terms of resemblance be grouped with any other words (some of them may be honorific forms). The remaining 56 words can be grouped in three broad groups containing together 48 words and in three small groups containing two or three words each. The first of the small groups containing KAMI is of words all drawn from languages of the Malayan group. Of the large groups, there is a strong resemblance between that starting ME and that starting with N and NE and the two groups probably could reasonably be combined into one larger group containing 30 words. The strong resemblance of the group of words beginning ME to the parallel group of words for the 1st person singular, also beginning with ME, is to be noted. As regards the group beginning WE, W is a rare letter for other pronouns and the number of words for WE involving it is notable, including even rather remote forms such as WOMEN (Chinese), WAXAI (Guahiban), WATASHI-TACHI (Japanese). Though less striking than for the pronouns earlier considered, the concentration of resemblances between words for WE is still remarkable.
The following are examples of particular interest:
WE English ME Finnish N Egyptian
WES Hittite MES Lithuanian NE Albanian
WIA Arawak Ml Hungarian NI Bulgarian
MY Russian NI Welsh
BIZ Turkish MU Hausa NOUS French
BITA Mongolian MA Persian ANOU Hebrew
MA Samoan NUNA Aranda
GU Basque MEMU Telegu NIHNA Arabic
KOU Tibetan
Of these examples, though geographically close, Finnish and Lithuanian are unrelated languages; the resemblance of Finnish and Telegu and the resemblance of Samoan and Persian are also striking. Of the words beginning with N, the set of words from French, Hebrew, Aranda and Arabic show a remarkable similarity. The degree of resemblance overall between words for WE goes beyond what is conceivable as a result of pure chance.
HE Though a good number of languages do not have a word for the 3rd person singular separate from the demonstratives, the words in the list can be grouped on the basis of resemblance as follows:
I HE SE ER TER AVAN EF
I HE ES ERA TERE AVAN
YI HIJ SU PONI
IA HU SIYA KARE T'A ATADU
IA HOU KIRI TOE TINI
IA THU
IA HA KO
AY HAN KOW
YAH HAN
IL HAN
EL HAN
EL HANN
ELE AN
ILLE ON
ITLE ON
LI ON
EGLI ON
E NO
ENE 0
YENA U
YEYE
Only some 4 words in the list appear to be isolated: EF PONI TINI ATADU are respectively Welsh, Guahiban, Bengali and Telegu. The remaining words are formed into two large groups, between them containing 41 words, and several small groups with 2-4 words each. There are resemblances between the two main groups e.g. I HE YI HIJ ENE AN YAH HA
but these are not extensive enough to form them into a single group. The small group starting with SE also shows some resemblance to the group beginning with HE: HE SE HU SU HIJ SIYA
but again it does not seem possible to incorporate it in the larger group. Nevertheless, the extent of resemblances between unrelated languages is again noteworthy as the following examples show:
I Manchu ENE Sumerian SE Irish
YI Hausa YENA Zulu SU Egyptian
IA Hawaian YEYE Swahili ES Georgian
AY Albanian SIYA Tagalog
YAH Hindi HE English
HU Arabic ER German
ILLE Latin HA Other Amerindian ERA Aranda
ITLE Eskimo HAN Swedish
LI Arawak AN Pahlavi KARE Korean
IL French KIRI Japanese
ON Yoruba
KO Tibetan ON Russian TER Nubian
KOW Thai NO Vietnamese TERE Mongolian
Of these examples, Manchu, Hausa and Hawaian have no known relation to each other nor obviously do Latin, Eskimo and Arawak. The resemblance of Tibetan and Thai may be due to language relation as also that of Swahili and Zulu but not Sumerian. Pahlavi is an Indo-European language but the resemblance between English, Arabic and one of the Amerindian languages obviously owes nothing to relationship, nor can the resemblance of Russian, Yoruba and Vietnamese. The resemblance of German and the Australian aboriginal language is striking as well as that of Nubian and Mongolian. There may be some borrowing between the Korean and the Japanese though the two languages are unrelated. There is obviously no relation between Irish, Egyptian, Georgian and Tagalog.
These results, not only the curious resemblances but the tendency for the different languages to use a limited range of sounds for the 3rd person singular, are in line with those for the other personal pronouns.
IT Many languages do not distinguish the pronoun IT from the demonstrative THAT or possibly from the pronoun HE. The result is that there are relatively few distinct words for IT. These can be grouped in terms of resemblance as follows:
YIH IT ONU LE ES KUGOT
YA ITO ONO LO SE
IA ITLE KON LO IRORO
IHA ISTO ON SORE
HA ID NO
IDAM NA ERA
ADI TER
DIA
DET
DET
DET
HET
SET
HIT
ETO
TO
TA
ATE
OTO
AUTO
ATU
ATHU
THU
Two of the words, KUGOT and IRORO, are completely unlike any of the others. They are from Korean and Arawak respectively. The remaining words form one large group where there is a broad resemblance, containing 23 words, one small group with 6 words, another with 5 words and a few small groups with 2 or 3 words each. The large group produces resemblances between remote languages which are of interest:
IT English DET Swedish
ITO Tagalog SET Egyptian
ITLE Eskimo HET Dutch
ISTO Portuguese
ID Latin
IDAM Sanskrit
ETO Russian
T'A Chinese
ATE Albanian
ATU Malayalam
AUTO Greek
OTO Hebrew
In the smaller groups resemblances of interest are:
ONU Turkish
NO Vietnamese
NA Samoan
ONO Czech
KON Tibetan
Some of the individual resemblances are striking and the
concentration of words for IT in the large group beginning with
IT fits the same pattern as has been found for other pronouns.
THEY In the same way as for HE and IT, a number of languages
make no clear distinction between the pronoun and demonstrative
forms. The words appearing in the Table for THEY can be arranged
in terms of resemblance as follows:
TE HE El U ABE ONI LAKOU
DE HEI ELES SU AVAR ONI
DE HEME ELLOS SIE VARU ONY KUDUL
DE HIE ILS ZIJ AVANGAL ONLAR
TEDE HINNI ILLI ESSI OK
THEY ITLAIT ESENI AWESAN WY
THUDO SEN ISHAN WAO KARERA
THEIR LA SILA AWOU
TAHARA MEREKA
TIR ENENE SIAD
ATA BONA
T'AMEN
ETNA POMONE
DA
CHUNGNO
VE
CE
Out of the 58 words, at least 8 have no clear relation to any other words (those in the last column); these are drawn from Hawaiian1 Korean, Hungarian, Japanese, Malay, Zulu, Guahiban, and Vietnamese respectively. The other words are arranged in groups of more nearly equal size than was the case for the other pronouns. The largest broad group contains 14 words but the group beginning TE could plausibly be associated with the group beginning HE which would make a larger group containing 19 words. The next largest group, beginning with U has 9 words.
Some interesting resemblances from the largest group are as follows:
TE Bulgarian THEY English
DE Swedish TAHARA Bengali
TEDE Mongolian THEIR Icelandic
DA Tucanoan TIR Nubian
ATA Albanian
Whilst Swedish DE is related to English THEY and probably to Bulgarian TE, all being Indo-European languages, the comparison with Mongolian and Tucanoan is interesting and of these with the Albanian. Bengali is also an Indo-European language included with the English and Icelandic for comparison with the Nubian form.
Examples from the other groups are:
HE Finnish ILLI Latin ABE Khoin
HIE Anglo-Saxon ITLAIT Eskimo AVAR Malayalam
HEI Arawak
HEME Hebrew ESSI Italian AWESAN Pahlavi
HINNI Arabic ESENI Georgian
SEN Egyptian ONY Russian
ONLAR Turkish
Words not mentioned in the examples which show more or less resemblance to each other mostly are from related languages e.g. ISHAN and AWESAN are Persian and Pahlavi. AVAR and AVANGAL are both Dravidian, WAO and AWOU are from related African languages.
Though for THEY there is not the same concentration in use of a very few sounds as has been seen for other pronouns, the number of resemblances between remote languages is still considerable
THIS The words for the demonstratives, such as THIS, can in the same way as the words for pronouns be grouped in terms of resemblance as follows:
E ZE SIS Dl I HIC LENEI
E EZ THIS Dl I HIKI LELI
ED SIS THES IDI IN HEEKE
EZ CECI THESSI DEZE IN WILl
ESTE CHE TOZI DIESER NII HAU
ESTE KEIA THEN DENNE INI HOUTOS XUA
ATE KETE TENTO DENNE ENE
ACEST ACEST DENNA EYINI AAHWN
ETOT QUESTO TIYAAHI INI YAH
VIDU AYAM BU
CAl-NAY ITO UNA
KONO NANA TAMA
WANNAN
There is not the same concentration of resemblances in one group as was seen for most of the pronouns; a wider range of forms seems to be used. There are five medium-sized groups, each containing 8 or 9 words and a number of small groups. Nevertheless, there are resemblances and links between some of the medium-sized groups; there is a broad resemblance between words in the group beginning SIS and the group beginning ZE and similarly between the group beginning SIS and the group beginning Dl. Some words in the two groups are obvious variants e.g. DEZE THESE, DENNE THEN. Five words appear to be isolated and to have no apparent resemblance to any other words in the collection. These words - WILl XUA HWN BU TAMA - are Mayan, Guahiban, Welsh, Turkish and Finnish respectively.
There are, however, a good number of interesting resemblances across language-families. Some examples are:
E Bengali ZE Hebrew SIS Lithuanian
E Sumerian EZ Hungarian THIS English
EZ Hungarian CHE Chinese TOZI Bulgarian
ED Pahlavi KEIA Hawaian
KETE Albanian THEN Egyptian
Dl Tibetan CECI French DENNE Norwegian
IDI Tamil ACEST Rumanian TENTO Czech
Dl Burmese
I Korean IN Nubian
HIC Latin I Malayalam IN Persian
HIKI Swahili INI Javanese
HEEKE Other Amerindian NII Thai
ENE Mongolian
EYINI Yoruba
UNA Eskimo
NANA Aranda
WANNAN Hausa
Some of the resemblances between distant languages are curious e.g. the Latin HIC and Swahili HIKI, the Egyptian THEN and the Norwegian (and other Nordic) DENNE. Some of the resemblances may be due to distant language relation e.g. that of the Lithuanian, Bulgarian and English, SIS THIS TOZI, but overall the conclusion is that resemblances across languages are at least as striking for THIS as for the pronouns.
THAT The words for THAT in the different languages can be grouped in terms of resemblance as follows:
A AZ SISE DA NE SORE ONZI
A ESE ISTE DE NA HORI
Al CELA ESSE DIE AN PF
DA KELA ESE DEN NAN
ATE QUELLA ASAU DEN NAN EKEINOS
THAT ACELA AZ DEN-DAR
THAET CAI-KIA IS MAN JOVAH
THETTA KEH ANIS ADI SAN
TOT KU ANAS VADU ANIS YAON
TAMTEN ANAS
TEN LELA ITU IYEN SU
TAUNA ITU HENE
TANA JENER LOKHO
TERE TUO HWNNW
BAHARA
IMA
HADAK
The pattern of resemblances is very similar to that found for words for THIS i.e. five medium-sized groups and a number of smaller groups. However, more of the words, 10 in all, have no apparent resemblance to any other words; these words, in the final column, are from Bulgarian, Egyptian, Greek, Hindi, Turkish, Zulu, Guahiban, Other Amerindian and Hebrew respectively (many of them are clearly related to the corresponding words for THIS in the same language e.g. Hebrew HADA HADAK, Turkish BU SU).
There are some resemblances, though not very extensive, between the larger groups. The group containing THAT has some parallels with the group containing DA DEN (mostly due to language-relationships).
The more interesting resemblances between languages are:
A Khoin AZ Hungarian ACELA Rumanian
A Malayalam ESE Spanish CELA French
SISE Tucanoan KELA Hawaiian
ATE Albanian ISTE Latin KEH Persian
THAT English ASAU Sanskrit CAI-KIA Vietnamese
TOT Russian
TEN Czech ANIS Hittite NA Chinese
TAUNA Eskimo ANAS Lithuanian NAN Hausa
TANA Aranda NAN Thai
AN Pahlavi
ANIS Hittite
ANAS Lithuanian
Of the resemblances between distant languages perhaps the most curious are those between Aranda (Australian aboriginal) and Eskimo, between Chinese, Hausa and Thai and between the Persian, Hawaiian, French and Rumanian
The degree of resemblances across languages is much the same as that found for words for THIS
THE Many languages make no use of the definite article or have no word for it distinct from a demonstrative THAT, or THIS The Table has words for THE in rather less than half the languages listed These words for THE can however be arranged in groups in terms of resemblance in the same way as the pronouns and demonstratives:
AY SE ERA -EN PA HO TO KA
-A JE N O
A DE HINN -BI
THE LE AN
DER IL -EN CIG
Y DEN EL ANG
LE NA WA
The words used for THE are more scattered in terms of resemblance than those for the pronouns and demonstratives The position is complicated where a language has differing forms for gender e.g. Greek HO HE TO. Examples of resemblance across languages which are interesting are:
-EN Danish
N Hausa
AN Irish
ANG Tagalog
NA Yoruba
HINN Icelandic
IS A good number of languages do not distinguish the verb
TO
BE Nevertheless, where IS exists as a separate word, there are
some resemblances which can be grouped as follows:
I AY JE UA ER DES BI
IS HAl CE AU ER ADA
IZAN CHY YW AR DIR BAINA
IST WA
EST NI
ESTA
ESTI IRUKU
EST
ESHTE UNNADU
ASTI UNTU
ESO
E ON
SHI
SHIH
E
AST
AST
The largest group is based on languages identified as
belonging to the Indo-European group. However, some of the
resemblances are interesting:
IS English UA Samoan
IZAN Basque WA Yoruba
ESO Khoin AU Egyptian
ESHTE Albanian YW Welsh
SHIH Chinese
CHY Thai
CE Hausa
Perhaps the most curious is the resemblance between the Basque and Indo-European words.
Before summarising and drawing conclusions from the examination made of resemblances between the words in many languages for pronouns and demonstratives, there is one other feature of the material which is striking, that is the extent to which there are resemblances not simply between words for the same pronoun or demonstrative but between the words for different pronouns and demonstratives. Rather in the same way as study of words for family relationships (MOTHER FATHER UNCLE AUNT) in diverse languages shows a high degree of resemblance but at the same time a certain fluidity in use so that words which mean FATHER in one language may mean UNCLE or MOTHER in another, so it appears that there is a certain fluidity between languages in using the same word-sounds for differing pronouns or demonstratives. This can be shown most clearly by the following comparisons:
Al means 'I' in Nubian TE means 'THOU' in Hungarian
AY means HE' in Albanian TE means 'YOU' in Finnish
Al means 'THAT' in Bengali TE means 'THEY' in Bulgarian
ME means 'I' in Irish and Georgian
NI means 'THOU' in Chinese
ME means 'WE' in Finnish NI means 'YOU' in Swedish
MY means 'WE' in Russian NI means 'WE' in Bulgarian
and Welsh
MI means 'I' in Etruscan NI means 'IS' in Swahili
NA means 'I' in Korean WIJ means 'WE' in Dutch
NA means 'IT' in Samoan WY means 'YOU' in Polish
NA means 'THAT' in Chinese WEWE means 'THOU' in Swahili
NA means 'THE' in Yoruba WY means 'THEY' in Welsh
BI means 'I' in Mongolian VI means 'YOU' in Bulgarian
BI means 'YOU' in Tucanoan VI means 'WE' in Danish
AZ means 'I' in Pahlavi
AZ means 'THAT' in Hungarian
ON means 'HE' in Russian I means 'HE' in Manchu
ON means 'IS' in Finnish I means 'YOU' in Danish
I means 'THIS' in Korean
HAN means 'HE' in Danish I means 'THIS' in Malayalam
HAN means 'I' in Guahiban SU means 'HE' in Egyptian
ESTE means 'THIS' in Rumanian SU means 'THOU' in Greek
ESTI means 'IS' in Greek SU means 'THEY' in Hausa
E means 'THIS' in Sumerian A means 'THAT' in Khoin
E means 'IS' in Italian A means 'THAT' in Malayalam
E means 'THIS' in Bengali A means 'THE' in Bulgarian
E means 'HE' in Khoin
YEYE means 'HE' in Swahili
BUT - there is no word like IC or EG meaning 'I' amongst the other pronouns and demonstratives
- there is no single word like ME or MI (meaning 'I') among words for HE IT THEY
- apart from the example given, there are remarkably few words for other pronouns and demonstratives (other than for YOU plural) like the group TE TU TI meaning THOU'
- no other pronoun resembles the group HE (meaning 'HE') other than the Finnish for 'THEY'.
The picture then is a mixture of some fluidity of sound/meaning relation coupled with some specificity.
The conclusions drawn from this examination of words for pronouns and demonstratives in a wide range of languages are:
(a) Amongst words in the different languages for particular pronouns and demonstratives, there is in nearly every case a high degree of concentration of resemblances into one, two or three main groups of words. The resemblances are so extensive and cut across language families in such a remarkable way that neither borrowing nor chance nor common language-descent appear adequate as an explanation. There is an established tendency for a limited range of sounds to be predominantly used for a limited range of meanings;
(b) Besides the resemblances between words for any single pronoun or demonstrative, there are also resemblances between words for different pronouns and demonstratives The common element of meaning between pronouns and demonstratives, i.e. that they are all indicative, seems to carry with it a greater probability that certain sounds will be used for one or other pronoun or demonstrative;
(c) Often resemblances between remote languages are more striking than those between related languages. Two principles appear to be at work, the first being the familiar dialectal variation by which originally c9mmon forms belonging to a historically prior single-language community have diverged to form descendant languages and the second being some relation between languages which goes beyond language families and may be universal;
(d) Such a universal or near-universal relationship cannot have a cultural or conventional basis but must derive from the underlying nature of language and from constraints imposed by the physiological character of the speech-process;
(e) It is probable that the degree of resemblance between words for pronouns and demonstratives derives from the similarity of function of pronouns and demonstratives in all languages. All pronouns and demonstratives can readily be associated with simple indicative gestures. The scheme of relationships between sounds, meanings and gestures presented in Part I of this book offers a plausible and consistent explanation.
This section concludes with a quotation from Meillet, one of the outstanding students of comparative linguistics. In speaking about the criteria for establishing language-relationships he said:
"Les rapprochements qui ne s'extendent pas … plus de deux dialectes sont peu sûrs, sauf raisons particulières; car la ressemblance de deux mots exprimant le même sens dans deux langues différentes peut être due à une rencontre fortuite; c'est ainsi que l'anglais BAD n'est pas apparenté, même de loin, avec le persan BAD signifiant aussi 'mauvais'; mais ce serait un hasard surprenant si BAD signifiait 'mauvais' dans une troisième langue sans un rapport avec l'anglais ou le persan. La coincidence de trois langues non contigues suffit donc pratiquement à garantir le caractère "indo-européen" d'un mot. ".
By these standards, the significance of the resemblance of the words for pronouns and demonstratives is established many times over; forms e.g. for THOU which resemble each other are found in some forty languages - many of which are remote from each other. This cannot establish language-relationship - it establishes some other universal principle of language-formation on the lines indicated above.
The starting point for this section is the study by Berlin and Kay published under the title 'Basic Colour Terms, Their Universality and Evolution.' They set out to examine how true it was that the perception of colour by speakers of different languages was determined by the manner in which the particular language encoded colour perceptions into basic colour words (e.g. white, black, red), the contention of the Whorfian school being words for basic colours are strictly not translatable between languages because the system of words available in one language is totally arbitrary with respect to comparable encoding in other unrelated languages. However, on the basis of a linguistic and psycho-physical investigation of the traditional doctrine in ninety-eight languages of diverse language families, Berlin and Kay concluded that colour-words are properly translatable, that eleven precisely defined colours serve as the perceptual focal points of all the basic colour words in all the languages of the world i.e. that in relation to colours the existence of semantic universals is established - in the some way, as in the immediately preceding section, it has been argued that the pronouns and demonstratives constitute a set of words with universal meanings translatable across languages. Put simply their conclusion is that whatever word is used for the colour RED in different languages, the speakers mean by RED the same actual colour, the same perception. This means that the set of words for colours is prima facie suitable for testing uniformities between languages of the sounds and meanings of words.
A more detailed account of the analysis which can be made of the material collected by Berlin and Kay is to be published elsewhere. For the purposes of this section, the manner in which the material can be used to explore the extent to which there seem to be systematic relations between the sound and meaning of words going across many languages (on similar lines to the analyses of pronouns and demonstratives in the preceding section) is illustrated by consideration of the words found for the colour RED. 9 languages have no separate word for RED. The words in the remaining languages can be grouped in terms of apparent resemblance as follows:
RE KORE POGH LAULAU MAADO TANTANKIN MYNFU
RED KYIREY KPOU LUAL ADOM TUTUKA
RARA' KIRAN PEL KULA 'AHMAR TSUKU
(GARA) KOKO PALA KULA ADARO BLEMA
(BARA) (KYAMA) PULA CIPSwUKA
('BABARE) (KONDON) (ANPALUKTAK) 'AS (SUBILA) PEAT
RARA (EKUNDU) (PPALKAHTA) AZGARH SIVAPPU
ERERENG COH MAPULA HUNG AZUFU CUWEPPE LOTOR
(EYEYENGO) CAH 'ILP'ILP HUNG 'AT'E
ERYTHROS (CERVENO) PIROS BEG SHILOWA
ROJO (KRASNYY) PIRIR VEVE CABAC
'GA VERMELL NCHI ABANG UHIE
CARA AKA NI (OWANG)
ASERAH JA DO OTI LICHI MBWAKI NAAIDAT
MERAH KACE DIDE S-WIGI
MERAH (GODDIOUDO) MOTANE CHINANA ENYUKI
SHAH DENG NYIAN
The result is that there are three fairly large groups and several smaller groups with about 13 completely isolated words and others with more distant resemblances. Out of the words listed, 42 are shown in the three larger groups. Within these groups there Are some interesting resemblances between words drawn from unrelated languages. For the first group, the languages from which the words are drawn are as follows:
RE - Bulu - Congo-Kordofanian (Africa)
RED - English
RARA - Hanunoo (Philippines)
GARA - Batak (Sumatra)
BARA - Batak (Sumatra)
'BABARE - Urhobo - Congo-Kordofanian (Nigeria)
CARA - Tarascan (Mexico)
ASERAH - Yibir - Afro-Asiatic (Chad)
MERAH - Malay (Malaya)
ERERENG - Nasioi - South Bougainville (South Pacific)
EYEYENGO - Poto - Congo-Kordofanian (Nigeria)
For the second group, the words ore drawn from the following languages:
KORE - Arawak - Andean-Equatorial (Surinam)
KYIREY - Songhai - Nilo-Saharan (Mali)
KIRAN - Fitzroy River (Queensland)
KOKO - Tshi (West Africa)
COH - Tzotzil - Mayan (Mexico)
CAH - Tzeltal - Mayan (Mexico)
KONDON - Songhai - Nilo-Saharan (Mali)
EKUNDU - Swahili
KYAMA - Songhai - Nilo-Saharan (Mali)
For the third group the words are drawn from the following languages:
POGH - Toda - Dravidian (South India) KPOU - Mende - Congo-Kordofian (Sierra Leone) PEL - Chinook Jargon - Amerindian (NW America) PALA - Hopi - Uto-Aztecan (SW United States) PULA - Tagalog (Philippines) ANPALUKTAK - Eskimo (Canada) PPALKAHTA - Korean MAPULA - Bisoyan (Philippines) 'ILP'ILP - Nez Perce - Penutian (United States) PIROS - Hungarian PIRIR - Nandi - Nilo-Saharan (Ethiopia)
Apart from the similarities in the main groups of words for RED, the following more restricted resemblances can also be noted:
TANTANKIN - Pozo - Hokan (United States) TUTUKA - Arunta (Australia) LUAL - Dinka - Nilo-Saharan (Sudan) LAULAU - Tanna Island (New Guinea) MAADO - Dazo - Nilo-Saharan (Eastern Nigeria) ADOM - Hebrew 'AHMAR - Arabic
There are a number of features of the collection of words for RED to which attention can be drawn:
(1) The above average number of words beginning with R or in which the R sound is prominent; the above average number of words similarly including K and P; in comparison only 3 words for BLUE in Berlin and Kay's list begin with K, none begins with R and none begins with P; out of 103 words for WHITE, only 3 begin with R. Out of 100 words for BLACK only 1 begins with R, none begins with P.
(2) Some reasonably close resemblances between remote languages:
RE Bulu KORE Arawak KOKO Tshi
RED English KYIREY Songhai COH Tzotzil
RARA Hanunoo KIRAN Fitzroy River
KPOU Mende
PIROS Hungarian PEL Chinook POGH Toda
PIRIR Nandi PALA Hopi
PULA Tagalog TANTANKIN Pozo
MAADO Dazo TUTUKA Arunta
ADOM Hebrew CARA Tarascan
GARA Batak
A similar analysis can be made of words for the other main colours WHITE BLACK BLUE. The picture for the other colours is much the same as that seen for RED; an apparently above chance degree of resemblance between remote languages, a tendency for a few sound-elements to be particularly prominent for each of the colours and for the words to be readily grouped (except for a relatively small number of isolated words) in terms of broad resemblance A few examples of resemblances are:
For WHITE For BLACK
BOPU Ngombe (Congo) TUI Nandi (Ethiopia)
PUPU Tiv (Nigeria) TIYE Ixcatec (Mexico)
POPO Sierra Popoluca
YOPA Queensland KURO Japanese
EUPE Swahili GURU Fitzroy River
ERA Bedauye IK Tzotzil (Mexico)
URA Tarascan YIK Sierra Popoluca (Mexico)
BURA Fitzroy River ZIKO Nupe (Nigeria)
EBORR Masai (Sudan) MADO Somali (Chad)
MAITUM Bisoyan (Philippines)
LIKAI Western Apache
LAGAI Navaho UNMA Queensland
LAGTI Hanunoo(Philippines) NEMA Shona (Rhodesia)
LEUKOS Greek HMA Mazatec (Mexico)
LOKWE Bari (Sudan) MANARA Queensland
For BLUE
BILU Samal (Philippines)
BLUE English
NILA Malayalam
NILA Urdu
LAN Chinese
LHIL LANNA Zuni (Mexico)
There is room for argument what precisely the mathematical chances are that by a random process of selection of sounds (from the millions of combinations possible) exactly the same combination or a closely similar combination of sounds should be selected in more than one language for a particular one of the unlimited number of distinct percepts and concepts identified by the human mind. But to say that all the resemblances noted between unrelated languages are the result of chance alone is not plausible. If the resemblances between words for the same colour were due to chance, then one would expect to find the same degree of resemblance between words used in different languages for the different colours, but this is not the case.
If chance is ruled out as an adequate explanation, it is equally implausible to assume that the degree of resemblance observed can be explained as due to hitherto unrecognised genetic relationships between geographically widely separated languages. Nor is it likely that borrowing can be the explanation for any large number of similarities. Borrowing is not an acceptable explanation between languages where there has been no special political, cultural or geographical reason for borrowing to take place.
If these other explanations are ruled out, then what is left is some universal tendency - or near-universal tendency - for similar percepts, in this case colours, to be associated with similar sound-sequences forming words. This does not operate absolutely because in that case there would not be many different words for the same colour) but tends to increase the probability that certain patterns of words will be found for certain percepts. It implies that along with the semantic universals in language (which Berlin and Kay's work seems to establish) there are phonological universals, which govern the formation of words.
This conclusion is in harmony with the evidence presented in earlier sections of this Chapter. Whether for colours or for pronouns and demonstratives or for the common words for parts of the body &c dealt with earlier, the pattern of results has in each case been the same i.e. a greater degree of resemblance between the sounds of words used in different languages for the same meanings than can be explained in any conventional way. The cumulation of evidence from different directions goes to support the basic hypothesis presented in Part I, that there is a natural relation between the sounds and meanings of words.
The presentation and verification of the hypothesis considered in this book have concentrated very largely on a single language, English, even though there has been fairly considerable examination of resemblances in certain fields between diverse languages. The same technique for assessing the degree of sound/meaning relationship as has been applied to English can be applied to other languages. Purely as an illustration of how this might be done, the following consideration of the relation between sound and meaning in Basque is appended. This is only an outline of a much fuller study that would be necessary.
(1) Words in Basque appear to be expressive (in much the same way as words in English were found to be expressive). The following are some examples:
CUCKOO KUKU SPIT THU
MOO MU LICK MILIKATU
MIAOW MIAU SNEEZE URTZINTZ
CHIRP TCHIUKA SNIFFLE ZURRUPA
WHISTLE CHICHTU SNEER MURRIKA
HOWL UHURRITU GNASH KIRRISKA
TINKLE DINDATU GNAW KARRUSKATU
CRACK KRASKA GONG DUNDUN
CHEW CHEHATU THUNDER DURRUNDA
(2) Words in Basque can be arranged in groups where the sound and meaning appear to be systematically related (as was demonstrated in English for e.g. the group of words CLAP CLIP CLASP &c). Examples of this are:
CLUTCH - HARTU GRIP - HARTZE CLENCH - HERSTU GROPE - HAZTATU CLAMP - HORTZ GRASP - HATCHEMAN CLIP - HARRAPATU GRAPPLE - KRAKO CLASP - KORTCHET GRATE - KARRAKATU CLAP - KLASKATU GRIND - KIRRISKATU CRACK - KRASKA GRAZE - KARAMITCHA
(3) However, comparison of Basque with a wide variety of other languages shows few resemblances for common words (in the lists of high-frequency words considered earlier in this chapter). Basque seems to be a genuinely isolated language in terms of vocabulary. Though a few resemblances exist, the difference between Basque and other languages seems to be about as complete as that between Korean, another isolated language, and other languages.
(4) Even for pronouns, the difference between Basque and other languages is consider-able Apart from the isolated resemblance Basque NIK meaning 'I' with NUK Egyptian1 NUKA Arawak, most of the Basque pronouns are distinct.
(5) The same applies to the Basque words for colours. There are no words in the other languages considered which resemble the Basque words for WHITE BLACK and BLUE and resemblances to the Basque word for RED are not very close.
(6) For English in Part I, a system of sound/gesture relationships was established. The isolation of Basque suggests that the system devised for English cannot be applied as it stands to Basque - but this does not mean that no natural (physiological) relation exists in Basque between the sounds of words and the meanings they have.
(7) It is not possible to develop in this book the evidence for believing that the sound meaning relation (linked to a gestural system) is as full and systematic in Basque as in English - this would need a complete repetition of the material contained in Part I but in terms of the Basque language - but there is reason to believe that a parallel sound/gesture system can be constructed in which for every sound in Basque there is a corresponding gestural element, with the total system taking the form of a transformation or transposition in a regular way of the system explained in Part I for English. The photograph immediately following shows for sounds in what in Part I were described as the Main Consonantal and Main Vocalic Groups the relation between the associated gestures in English and Basque. Essentially, the proposition is that the English system is transposed into a different set of co-ordinates rather as a sphere could be systematically distorted into an ellipsoid so that any patterning on the surface of the sphere would undergo a regular change of relative position and shape (an idea similar to that of D'Arcy Thompson's method of transformation of biological forms through regular distortion of a reference set of co-ordinates) See the discussion of the relevance of this to language in Lenneberg's Biological Foundations of Language pp. 245 ff.
The same approach, as has briefly been described for Basque, could be followed for other languages. Many languages on a cursory examination appear to make use of a sound/gesture system very similar to that described for English, including in particular other Indo-European languages, the Dravidian languages, the Malay-Polynesian group, Aranda, Japanese and a good number of non-tonal Amerindian languages. Other languages, like Basque, appear to be based on a different sound/gesture scheme (though with a systematic transformation of form, on the lines briefly indicated above, from the English system). This appears to be the case for Korean, Finnish, Hungarian, the Semitic languages, and the Bantu languages. Chinese, as a tonal language, so far does not appear to fit this pattern but no doubt in time it may prove possible to relate it to the sound/meaning system found in non-tonal languages (see the difficulties recorded by Roger Brown and others in demonstrating a better than chance probability of guessing words between Chinese and non-tonal languages).
This account is inevitably extremely compressed and there is no possibility of setting out in the present book the evidence on which the suggestions made in the immediately preceding paragraph are based. There is obviously an immense field left for research There seems no reason to believe that the systematic relation of sound and meaning described in English is a peculiarity of English or of any single group of languages Further research will surely show that it is only one particular aspect of a general relation, operating in all languages, between sound and meaning, perception and gesture.
The work of exploration of the hypothesis has only just begun A wide field of enquiry remains to confirm or refute the ancient view that language is phusei and not thesei, that its basis is in human physiology and neurology and not in mere habit or convention. It may be appropriate to conclude with two quotations which bear on the basic issues underlying the hypothesis and material presented in this book, first how plausible it is to assume a structural relation between the sounds of language and meaning, between gesture and perception and secondly how far chance is a sufficient explanation for resemblances and apparent ordering of phenomena.
On the first point, Charles Darwin, the naturalist, quoted approvingly a French anatomist, Gratiolet, on the nature of human expression:
"II résulte de tous les faits que j'ai rappelés que les sens, l'imagination et la pensée elle-même, ne peuvent s'exercer sans éveiller un sentiment correlatif et que ce sentiment se traduit .... dans toutes les sphères des organes extérieurs suivant leur mode d'action propre comme si chacun d'eux avait été directement affecté".
The human individual functions as a totality. The modification in his system produced by thought or perception is accompanied by modification in the external processes which take the form of the patterning of articulation (producing speech-sounds) and the patterning of muscle-state (producing gesture and other physical expressive action).
As regards the mathematical probability of chance as the source of observed order, Charles Darwin's grandson, the physicist Sir Charles Darwin, described in his closing address to a Symposium (1957) on Observation and Interpretation in the Philosophy of Physics a relevant calculation he had made. Starting from the well-known idea that the whole of the Bible could be produced by a sufficiently prolonged random process, he said:
"I once amused myself by seeing how long it might be. I took an army of a million, starting 3000 million years ago when the earth first came into existence The first three words of the Bible are 'In the beginning' and I found that even after this immense time and with this immense number of typists, it was no more than an even chance that these three words would be found somewhere in the texts".
Language and Evolution: Books, Presentations and Papers
Motor Theory of Language