The First Congress of Ethnozoological NomenclatureUniversity of GeorgiaFolk names for fundamental ethnobiological categories have been shown to be governed by regular nomenclatural principles. Two principles at work in ethnozoological nomenclature – onomatopoeia and metaphorical description of some observable property of the organism – are fairly well established as the basis for naming many folk genera. A third but less understood principle is that associated with what has been c
Folk names for fundamental ethnobiological categories have been shown to be governed byregular nomenclatural principles. Two principles at work in ethnozoological nomenclature –onomatopoeia and metaphorical description of some observable property of the organism – arefairly well established as the basis for naming many folk genera. A third but less understoodprinciple is that associated with what has been called sound symbolism. In the languages oftraditional peoples, semantically opaque names for animals often exhibit sound-symbolicproperties that humans unconsciously recognize as capturing some aspect of the fundamentalessence or nature of the creature being named. How is this to be explained, in spite of thechanges that have taken place in human verbal communication since the beginnings of what one might call full-blown language? In what ways are these principles related to more generalprinciples of natural classification based on shape and movement? If verbal mimesis represents acritical stage in the evolution of human cognition, what informed speculations can be brought tobear on what might be called the First Congress of Ethnozoological Nomenclature?
By establishing a harmony between a thing and its name, we conform to psychic habit as old ashumanity (Vendryès [])
I take for granted, then, that there are some similarities between the experiences we have throughdifferent sense organs [and] that in primitive languages one finds much evidence for assuming thatthe names of things and events often originate according to this similarity between their propertiesin vision or touch, and certain sounds or acoustical wholes1 (Köhler : )
I have given my paper the somewhat through-the-looking-glass title ‘The First Con-gress of Ethnozoological Nomenclature’. It is appropriate, then, to begin with part of the conversation between Alice and the Gnat that I included as the epigraph in Ethnobiological classification (Berlin : i):
‘What’s the use of their having names,’ the Gnat said, ‘if they won’t answer to them?’ ‘No use to them,’said Alice, ‘but it’s useful to the people that name them, I suppose. If not, why do they have namesat all?’ (Carroll : ).
Given all that we have learned about the relationship of classification and nomencla-ture since Alice’s time, I think we can agree with her that names for creatures must be
The First Congress ofEthnozoological Nomenclature
useful ‘to the people that name them’. Much less understood, however, is just how andwhy people decide on the names that they bestow on the living creatures that surround them.
So, join me in a fanciful, but perhaps not so implausible, ‘First Congress of Ethno-zoological Nomenclature’, where we listen in on part of the discussion of its delegatesin one of their first workshops. The scene opens on a group of sages squatting around an evening campfire somewhere engaged in a serious discussion on a fundamentalnomenclatural question, one asked by Roger Brown now more than half a century ago:how shall a thing be called?2
As we come within earshot of our circle of naming specialists, one of the elders isarguing: ‘Then let us assume that if our goal is to propose a name that is a goodmnemonic for the referent to which it is assigned, surely an unambiguous descriptivephrase seems ideal’. ‘Of those with the enormous bills’, he continues, ‘the words “theone with underparts, wings and tail black, and upper and lower tail coverts yellow andcrimson” are certain to distinguish it from all others of its class’ (Fig. a).
‘Quite so’, agrees the elder to his right, ‘but such a principle seems hardly appro-priate for the inconspicuous grey one that is so common to the forest edges of ourgarden plots. It seems to call out its name every time it speaks’.
‘Of course you refer to the one that says: uweá, uweá, páipainch.áaa– páipainch.áaa– uweá, uweá, páipainch.áaa– páipainch.áaa. There seems little doubt but that pái-painch3 should be its name’ (Fig. b). A general nodding of heads all around indicatesunanimous assent.
‘And your principle would apply equally well to the slippery-slick-skinned ones,suákarep, karákarás, chirirí, pokarí, and warétete, that call their names in the night’, addsUncle Skinny as he puts another branch on the fire.
Figure 1. The two birds discussed at the First Congress of Ethnozoological Nomenclature ((a)Ramphastos vitellinus; (b) Lipaugus vociferans. Reproduced by permission of John A. Gwyne (a) andGuy Tudor (b) in Hilty 2003).
‘This works well enough when they lead us to their names by their calls’, pipes upone of the younger members in the circle, a nomenclatural rebel of sorts but who wassavvy enough to get invited due to his political connections. ‘But what of those thatdon’t call their names and whose descriptive designations are simply ponderous andboring?’ Then, looking as if he had discovered the key question, he asks: ‘If I were topropose wámpang [morpho butterfly] and wichíkip [small, inconspicuous butterfly]as the names for two of the silent flutter-flat-wings, is there any doubt about whichname we should apply to each?’ (Fig. ).
‘The young one speaks the truth’, responds the convener of the congress. ‘None theless’, he concludes, ‘I must confess that just why this should be so is not at all clear tome. For reasons beyond my explanation, it just seems appropriate and easy to namethem in this fashion’.
Our specialist in ethnozoological nomenclature makes a good point. We find it rel-atively simple to provide explanations for animal names based on descriptive phrasessuch as red-winged blackbird or redheaded woodpecker and we can easily offer folk ety-mologies for names such as catfish, armour head, or bigeye. Likewise, with little
Figure 2. The two kinds of butterfly discussed at the First Congress of EthnozoologicalNomenclature (wámpang large butterfly, top; wichíkip small butterflies, bottom (Butterfly Utopia n.d.,reproduced with permission).
conscious effort we can think of good reasons for the use of onomatopoeic bird namessuch as whip-poor-will and bobwhite. Much less understood, however, are those animalnames motivated by the principles that Hinton, Nichols, and Ohala have referred toas synaesthetic sound symbolism, ‘the process whereby certain vowels, consonants, andsuprasegmentals are chosen to consistently represent visual, tactile, proprioceptiveproperties of objects, such as size or shape’ (: ). Some of these distinctions are setout in Figure .
Early studies of phonaesthesiaOne of American linguistics’ major figures, Edward Sapir, noted that synaestheticsound symbolism is at work when we focus on ‘the expressively symbolic character ofsounds quite aside from what the words in which they occur mean in a referentialsense’ (: ). Roman Jakobson, principal founder of the Prague School of struc-tural linguistics and phonology, provides the beginnings of a psychological explana-tion for synaesthetic sound symbolism when he writes:
The intimacy of the connection between the sounds and the meaning of a word gives rise to a desireby speakers to … complement the signified by a rudimentary image. Owing to the neuropsychologi-cal laws of synaesthesia, phonetic oppositions can themselves evoke [sensations] of pointed androunded, of thin and thick, of light and heavy (Jakobson : ).4
Synaesthetic sound symbolism can be thought of as the cross-modal mapping thatunites specific speech sounds and one or more distinct sense modalities (sight, touch,smell, taste). Recently, a group of electronic vocal composers intrigued by the chal-lenge of making the ‘voice visible’ have captured the essence of the meaning of syn-aesthetic sound symbolism in a single term: phonaesthesia (Levin, Lieberman, Blonk& La Barbara ). This is the term I will use for the sound-symbolic but non-onomatopoeic representation of non-acoustic phenomena.
An overriding feature of the early experimental work on phonaesthesia has been itsfocus on abstract figures, manufactured objects, nonsense words, and antonyms innatural languages (for some classic early work see Brown a; b; Brown, Black,& Horowitz ; Brown & Nuttal ). Much less research has been carried out to
determine the extent to which phonaesthesia functions as an unconscious psychologi-cal motivation that might drive the naming of animals in natural languages. The preliminary data presented here are meant to support the claim that the role ofphonaesthesia in ethnozoological vocabulary is greater than any of us might haveexpected. Furthermore, recent work in neuroscience suggests that phonaesthesia mighthave played a significant role in the evolution of language, a topic to which I will returnat the close of the paper.
However, to place these data in their proper context, let me review two early exper-iments that set the stage for scores of studies of phonaesthesia for three-quarters of acentury. For most American linguistic anthropologists and psycholinguists, Sapir’sclassic ‘A study in phonetic symbolism’ paper is probably the best-known exper-imental work. In one of several studies, Sapir showed his subjects figures of two tablesof different sizes, indicating that the nonsense words mil and mal were the tables’names. He then asked subjects which name indicated the larger table. They invariablychose mil for the smaller table and mal for the larger (Sapir ).
Sapir’s research finds its origins in an earlier study by another great linguist, OttoJesperson, in his little-read paper titled ‘Symbolic value of the vowel i’5 (Jesperson
[]). Although the data discussed there were anecdotal, they led Jespersen to con-clude that in natural languages ‘the sound [i] comes to be easily associated with small[things] and [u o a] with bigger things’ ( []: ). Common examples of theprinciple of size-sound symbolism can be seen in the antonyms for ‘large’ and ‘small’in many languages of the world: for example, Spanish grande – chico, Tzeltal muk’ul –ch’in, Aguaruna muun – pipich, and Greek macros – micros (for further supporting data on size-symbolism, see Nichols ; Sapir ; Swadesh ; Tsur ; ;Ultan ).
A second experiment, even more widely known than those of Sapir, was carried outby the Georgian psychologist Dimitry Uznadze6 () and then elaborated five yearslater by Wolfgang Köhler, the father of Gestalt psychology (; ). Köhler askedhis subjects to observe the two drawings seen in Figure and to assign them the non-sense names takete and maluma or the basis of what name seemed most appropriatefor each figure. More than per cent of the subjects selected takete for the angulardrawing and maluma for the circular one.
Köhler’s experiment has been replicated many times (see Fox ; Irwin & Newland ; Lindauer ; for some early examples), and at least once
cross-linguistically among Lilongwe-speaking children of Lake Tanganyika (Davis). I have also informally conducted the experiment regularly with students in myevolution of human cognition classes, always with similar results.
Extending Köhler’s experiments on phonaesthesiaRecently, I conceived a modified version of the Köhler study. I hoped to determinewhat subjects might do if presented with the two figures and asked to produce theirown names for them? Would the same principles of phonaesthesia that appear tounderlie the association of maluma and takete, with their corresponding figures, beinvented de novo?7 Here, briefly, are the details of the study.
Twenty-two students in two introductory philosophy classes at the University ofGeorgia8 were requested to be of assistance to Steven Spielberg in his plans to inventa new language called Droidese. The language would be spoken by cartoon charactersin a new animated film to be titled Return of the Droids. The sound system of Droideseincluded the consonant and vowel phonemes seen in Table .
Students were also informed that a standard rule of Droidese phonology requiresthat the consonants and vowels of Droidese words take the canonical shape CVCVCV,for example dawelu. Given these rules for the form of Droidese words, I then askedthem to invent names that they felt would be most appropriate for Köhler’s figures(subjects were not told of the maluma takete experiment).
The results of the simple study show that vowels and certain consonants of Droidesewords exhibit some interesting features (Table ). The sounds selected to form theinvented names for the maluma and takete figures show significantly different patternsof distribution – maluma-like names prefer back vowels [u] and [o] while takete-liketerms favour the front vowels [i] and [e] (Table a).
For consonants, voiced stops were predominantly selected in the formation ofwords for the maluma figure (takete terms were about equally divided on the value ofvoicing) (Table b). The patterned distributions of these vowels and consonants wereboth statistically significant. Examples of some contrasting invented names are seen inTable c.
Some typical reasons provided by students to explain their motivations for theirinvented names were:
‘Maluma’ terms: sounds seemed to flow smoothly, words sound bubbly, rolling, circular, lazy-sounding vowels, round shape, loose, soft, slow.
One student gave the answer, ‘For reasons beyond my explanation, it just seemedappropriate and easy to name them as [I did]’, a comment, you’ll recall, similar to the one provided by the convener of our First Congress of Ethnozoological Nomenclature.
Students’ responses focus on two kinds of visual sense impressions evoked by thesounds of their invented names: one is shape (angular, sharp, curved, jagged-edged,pointy, short, long, diagonal, rigid, twisted); the other is movement (rapid, piercing,jamming, flowing smoothly, bubbly, rolling, fluid, slowly).
These data show that phonaesthesia is associated in significant ways with the sensorysensations of size (Sapir’s mil, mal), and shape and movement (invented words formaluma, takete). While an adequate psycholinguistic explanation for these results hasyet to be proposed, recent research has gone far in providing us with partial answers
as to why phonaesthesia seems to be so intuitively natural. Building on earlier work ofJohn Ohala (), Hinton, Nichols, and Ohala provide the following synthesis:
[H]igh tones, vowels with high second formants [F2] (notably /i/), [a vowel’s formants are acousticproperties] and high-frequency consonants are associated with high-frequency sounds, small size,sharpness, and rapid movement; low tone, vowels with low second formants [F2], (notably /u/ [and,we could probably add, /a/ and /o/], and low-frequency consonants are associated with low-frequencysounds, large size, softness, and heavy, slow movements (: ).9
These perceptual properties of phonaesthesia can be illustrated in Table .10
Phonaesthesia in ethnozoological nomenclatureIn what ways might these non-arbitrary sound-meaning associations of movement,size, and shape, so typical of phonaesthesia, be applicable to how animals are namedby peoples in traditional societies? In earlier research, I described how size-soundphonaesthesia could be shown to be at work in the ethnozoological nomenclature forbirds and fish among the Huambisa, a Jívaro-speaking people of Amazonas, Peru(Berlin ; ; see also comparable work on Malay fish names, Berlin ).Phonaesthesia was prevalent in both lexical domains – smaller birds and fish tendedto exhibit names formed with the high-front vowel [i] while larger birds and fishfavoured vowels [u] and [a] significantly greater than expected by chance. In addition,but only mentioned in passing in my first reports, it appears that fish and bird namesdiffer significantly in the distribution of high acoustic frequency consonants: birdnames overwhelmingly prefer initial voiceless stops [p, t, k] and voiceless affricates [ts, ch] before the high-frequency vowel [i]; when these consonants occur as initialsegments in names of fish, they overwhelmingly co-occur with the low-frequencyvowels [a] and [u].
In addition, fish names show a distinct preference for nasal consonants, sound seg-ments with extremely low sound frequencies. Thus, relative size can be said to bemarked by vowel quality whereas movement appears to be partially marked by acousticquality of consonants. These data also suggest that a combinatorial effect is achievedwhen the vowel and consonant parameters act in concert.
Might the names of other groups of animals show a similar sound-symbolic namingbias, and, if so, how widespread might the bias be across other languages? I began to
Table 3. Some perceptual properties of phonaesthesia.
Semantic dimension of phonaesthesia
Sound segment Movement Shape Size Sound
Front vowels, Rapid, quick, fast, Long, straight, Small High tonehigh-frequency abrupt, short, agile jagged, skinny, sharp,consonants thin, slender, angular,
lanky, pointed, flat (a special case of long)
Back vowels, Slow, sluggish, Spherical, short, Large Low toneLow-frequency lumbering, flowing, stocky, fat, round, consonants waddling, awkward, solid, hefty, squat,
smooth rotund, smooth
collect data on this question when I was struck by the natural, non-arbitrary natureof the Aguaruna names for the lumbering South American tapir, pamáu, and the quickand agile squirrel, wichíng (compare wámpang and wichíkip in Figure ; see also Berlin).
My loosely formed hypothesis was that if the principles of phonaesthesia were atwork, one should expect to find terms for tapir comprised of sounds of low acousticfrequency (large size, slowness), and words for squirrel marked by sounds of highacoustic frequency (small size, quickness). I gathered comparable data from a sampleof twenty-five South American Indian languages each from a distinct linguistic family.The results of this comparison are seen in Table . The data show that the high-frontvowel [i] and the high-frequency voiceless stops [t] and [k] mark terms for squirrel
Table 4. The results of a comparison of high and low acoustic frequency sounds in words for squirrel and tapir in twenty-five languages.
a)
vowel [i] Squirrel Tapir Total
+
−
total
two-sided P value is .
b)
vowel [a] Squirrel Tapir Total
+
−
total
P = .
c)
Stops [t], [k] Squirrel Tapir Total
+
−
total
two-sided P value is .
Labials [p], [b], [m], [w] Squirrel Tapir Total
+
−
total
P = . (not quite significant)
(small size, rapid movement?) while the sounds [i], [t], and [k] are not favoured in thenames for tapir (due to their large size and slow movement?).
Are there naming contrasts relating to other creatures that might form the begin-nings of a database on which we might base a theoretical explanation of phonaesthe-sia in ethnozoological nomenclature? Turning again to South America, I selected twogroups of birds, the rails and the tinamous, both of which differed significantly inshape (and movement?). Drawings of a species from each group are seen in Figure .Visually, these two birds might be thought of as metaphorically analogous to the taketeand maluma figures in Köhler’s studies. As a thought experiment, imagine that thenames of the two creatures in Figure in some indigenous language spoken in theUpper Amazon of Colombia were maluma and takete. Which name would one assignto each figure? In another experiment, I aimed to hold my own First Congress ofEthnozoological Nomenclature.
Using the instructions developed for Köhler’s maluma and takete figures, I sub-stituted the takete/maluma figures with drawings of a rail and a tinamou. My hypothesis was similar to the first study: on the basis of visual clues based on form(and the real or metaphorical association of form with type or speed of movement),names for the rail should show high-frequency consonants and vowels while thosechosen for the tinamou would be of low frequency.
I asked sixty-two undergraduate anthropology students,11 none of whom had par-ticipated in the earlier experiments, to invent names for the two birds. I passed aroundtest sheets with the sound system of Droidese accompanied by drawings of each bird.I further explained Droidese word formation as outlined in the first experiment butstated that all words must take the canonical form CVCVC, a highly common conso-nant-vowel sequence for word formation in many of the world’s languages.12
When students had finished writing their invented names for each of the birds, Ithen asked them to watch me write two words on a white board at the front of the
Figure 5. Aramides cajanea (rail) (a) and Tinamous major (tinamou) (b) (reproduced with thepermission of Guy Tudor from Meyer de Schauensee & Phelps 1978).
class, stating that if the words I had written might be assigned to the rail and thetinamou, which name would go with what bird? The names were taket and malum,variants of Köhler’s terms. Students then wrote each of these names next to the birdon their test sheet to which they felt it most appropriately applied.
First, as you would have predicted, sixty-one of the sixty-six students ( per cent)stated that taket was the most appropriate name for the rail. When we look at theinvented names for rail and tinamous, we see a familiar pattern: high-front vowels [i]and [e] are preferred for the rail, back vowels [o] and [u] for the tinamous (Table a).With consonants, we see that voiceless stops [p], [t], and [k] are favoured in theinvented names for rail, much less so for tinamou (Table b).
The reasons students gave for their naming responses support the principles ofphonaesthesia for shape and movement discussed earlier:
What are the naming patterns for rail and tinamou in the languages of the tradi-tional societies who actually know these birds intimately? With the goal of gatheringinitial data on this question, I focused once more on South America. I soon discov-ered that gathering lexical data for these bird species proved more difficult (andperhaps less reliable) than my earlier efforts to develop a comparative list of terms fortapir and squirrel. It is common that dictionaries or word lists do not indicate to whichspecies a particular term may apply. When dictionaries do list terms for tinamou andrail, names for tinamous, the most obvious and culturally significant group, are citedmore often than terms for rails. Furthermore, a folk name might include several bio-
logical species, some of the same scientific genus, some not. The ‘true’ rail and tinamouspecies in South America are:
Rails (Rallidae): Aramides cajanea, A. axillaries, A. erythroptera, Rallus longirostris.
Tinamous (Tinamidae): Tinamus major, T. tao, T. guttatus, Crypturellus soui, C. cinerus, C. undula-tus, C. varigatus.
As I believed it important to draw a sample of languages that included both a wordfor ‘rail’ and ‘tinamou’, I have limited my sample to languages where I am fairly con-fident of the terms listed for the large tinamous, Tinamus spp. (mostly commonly, T.major and T. tao) and species of the Wood Rail, Aramides spp. (most commonly, theGrey-Necked Wood Rail, Aramides cajenea). This resulted in a sample of seventeen lan-guages of distinct language families, or, when in the same family, showing rail andtinamou terms that are not obvious cognates, as is the case with Sataré and Urubu ofthe Tupian language family (Table ).
Conclusions and possible explanationsEven with such a small sample, if phonaesthesia is at work here, the principles notedfor shape and movement should be evident. The distributional patterns of high- versuslow-frequency vowels, it turns out, are not significant but the predicted patterns forconsonant phonaesthesia are suggestive. An angular, sharp, long-legged, streamlined-bodied rail ought to show a preference for voiceless consonants, especially voicelessstops, while the rounded, short-legged tinamou should not favour these sounds. Tablea shows that high-intensity stops [t] or [k] occur in all terms for rail; [p], the leastintense of the stops, occurs in just two names.
In contrast, nasals are consonants with a particularly low acoustic frequency andshould connote slow, round, plump, fat, soft, squat, heavy creatures such as thetinamou. These sounds should be avoided in terms for rail and favoured for tinamous(Table b).
Finally, all labial consonants, [p], [b], [m], and [w] exhibit low frequency in termsof air flow and as a consequence should be associated with the expected principles ofphonaesthesia governing shape and movement. Terms for rail should show signifi-cantly fewer labials than do terms for tinamous (Table c).
The distributional patterns observed for consonant phonaesthesia in these seven-teen languages are strong. It is likely that data from additional languages will confirmthe expected principles of shape/movement sound symbolism. This expectation isstrengthened given the studies described earlier in this paper.
In searching for an explanation of these findings, the data show that the physicalproperties (relative acoustic frequencies) of vowels and consonants play a major role.The high-front vowel [i] has an acoustic frequency of around - Hz as itssecond formant (F2, in acoustic terminology) while the back vowels [u, o] exhibit F2
between and Hz (Fig. ).Physical acoustic parameters are also of relevance in consonant phonaesthesia.
Voiceless stops are characterized by an ‘explosion’ in the normal airflow during artic-ulation since in ‘consonants, voiceless obstruents have higher frequency than voicedbecause of the higher velocity of the airflow, … dental, alveolar, palatal and front velars[have] higher frequencies (of bursts, frication noise, and/or formant transitions) thanlabials and back velars’ (Ohala : ).
Quechua Quechuan Aramides sp. KutsraTinamus major Yutu
Sateré-Mawé Tupian Aramides cajanea tarangkuTinamus major urit’iwato
Siona Tucanoan Aramides cajanea bo’teTinamus tao anka
Ulwa Misumalpan Aramides cajanea kudahTinamus major wankamara
Urubu Tupian Aramides cajanea sarakurTinamus major inambu
* ii = long vowel.Language sources: Achuar: Fast, Warkentin & Fast 1996, Apalai: Jensen 1998; Arabela: Rich 1999;Bora: Thiesen & Thiesen 1998; Capanahua: Loos & Loos 2003; Chayahuita: Hart 1988; Cuiba:Ortiz & Queixalos n.d.; Ese-Eje: SAILDP n.d.; Jarawara: Vogel, n.d.; Kandoshi: Tuggy 1966; Piaroa:E. Zent, personal communication; Mashco-Piro: SAILDP n.d.; Quechua: Irvine 1987; Sateré-Mawé:Jensen 1998; Siona: Wheeler 1987; Ulwa: Dictionary of the Ulwa language n.d.; Urubu: Jensen1998.
Furthermore, the voiceless stops [p], [t], and [k] are graded in terms of the inten-sity of the airflow associated with their production (p < t < k) (Fig. ). Ohala states itas follows:
[B]oth [t] and [k] have high-intensity bursts. One factor is that they are inside the vocal tract andthus exploit the downstream resonances. [p] can’t do this since there is no downstream cavity. [t]has a shorter cavity than [k] but it has the advantage that the air stream of the release can hit theteeth (lower and upper incisors) and these act as what the aerodynamic people call ‘spoilers’ or baffles,creating another source of turbulence (Ohala, personal communication; see also Tsur ).
A second parameter at play in phonaestheia is the articulation of speech sounds,something that might be referred to as vocal mimesis, the unconscious use of mouthgestures to indicate metaphorically the inherent qualities of the thing being named.This process was first systematically discussed by the British speech therapist SirRichard Paget in his Human speech (), published just a year after Sapir’s soundsymbolism study. Paget argued that:
The sound of the word is frequently found to be due to postures and gestures of the organs of artic-ulation which bear a pantomimic relation to the idea or action to which the word refers. From this
Table 7. Consonant phonaesthesia in names for rails and tinamous in seventeen languages.
(a)
[p], [t], [k] Rail Tinamou Total
+ 17 10 27− 0 7 7
Total 17 17 34
p = .0072.
(b)
Nasals Rail Tinamou Total
+ 3 10 13− 14 7 21
Total 17 17 34
p = .0324.
(c)
Labials Rail Tinamou Total
+ 6 14 20− 11 3 14
Total 17 17 34
p = .0134.
we infer that human speech arose out of a generalized unconscious pantomimic gesture language(including the tongue and lips). The gestures [of speech] were recognized by the hearer because thehearer unconsciously reproduced in his mind the actual gesture which had produced the sound (:).13
Continuing,
… [i] is made by pushing the tongue forward and upward so as to make the smallest cavity betweenthe tongue and the lips while o-o or aw-aw [o] and [a] are the results of a lowered tongue, produc-ing a large mouth cavity (: ) (Fig. ).
Paget’s views mesh closely with Merlin Donald’s speculations in Origins of themodern mind: three stages in the evolution of culture and cognition (Donald ; ;also see ) and the evolutionary speculations of Peter MacNeilage and his colleagueB.L. Davis on motor mechanisms in speech production (MacNeilage ; ;MacNeilage & Davis ; ). In speculating on the role of mimesis in the devel-opment of human language, Donald notes that
the principle of self-triggered voluntary retrieval of representations had to be established in the brainbefore the highly complex motor acts of speech would have been possible. Phonetic skill [makingspeech possible] depends on the basically mimetic ability of individuals to create rehearsable andretrievable vocal acts, usually in close connection with other mimetic acts. In a word, language … islayered on top of a mimetically-skilled phonological system (: ).
Figure 6. A spectrogram of the words heed, hid, head, had, hod, hawed, hood, who’d as spoken by a male speaker of American English. The locations of the first three formats are shown by arrows.(after Ladefoged n.d., reproduced with permission.)
Donald is at a loss, however, to provide a plausible theory for the development oflexicon in the first place. He states that, granted phonological agility, ‘Lexical inventionis not yet understood in terms of mechanism. There is no viable computational model ofthis process’ (: , emphasis added). None the less, Donald does anticipate sound-symbolic processes in early word formation when he continues: ‘Lexical invention mustwork according to a metaphorical principle’ (: ; see also Foster ; Hiraga; Lakoff & Johnson ). If phonaesthesia is, above all, metaphorical, then soundsymbolism would be a good candidate to first drive lexical representation in spokenlanguage.
These speculations also tend to support Mithen’s () proposal that the humancapacity for metaphorical thought arises from increasingly complex human symbolicprocesses as a result of what he calls ‘cognitive fluidity’ across the four principaldomains of intuitive knowledge – social, natural history, technological, and linguistic.The cross-modal metaphorical properties of phonaesthesia would serve only to facilitate the cognitive fluidity that Mithen suggests is essential for higher cognitivedevelopment (see Mithen , esp. chap. , and in this volume). In this light, thephonaesthesia that we find in ethnozoological vocabulary could have actively con-tributed to the development of humans’ natural history intelligence in ways that aresupremely intuitive – and natural.
Recently, and unknown to me until this paper had already appeared in draft form,these plausible yet speculative ideas have received some tentative neurological support
Figure 8. Relative positions of the tongue in the production of sounds [i] (as in heed) and [a] (as inhod) (after Ladefoged 2002: 115, reproduced with permission).
from the research of cognitive scientists S.V. Ramachandran and E.M. Hubbard ().They agree that what we have been talking about is ‘a kind of sensory-to-motor synaes-thesia which may have played a pivotal role in the evolution of language’ (: ).They argue that ‘the reason [subjects get the names right in the maluma/takete exper-iment] is that the sharp changes in visual direction of the lines mimics the sharp [pho-netic] inflections of the sounds as well as the [mimetic] movements of the tongue onthe palate’ (: ).
They conclude that ‘the representation of … lip and tongue movements in [our]motor brain maps may be mapped in non-arbitrary ways to certain sound inflectionsin auditory regions [of the brain] and the latter, in turn, function as non-arbitrarylinks to an external object’s visual appearance’ (: ).
We may tentatively conclude that non-arbitrary sound-symbolic, phono-mimeticreference must have had enormous adaptive significance for our hominid ancestors asthey began to play the naming game in earnest. It is not too great a stretch of the imag-ination to suggest that the intuitively plausible and metaphorically motivated princi-ples of phonaesthesia served to drive the development of the lexicon in general. Thestrength of this type of sound symbolism is manifest in our names for the animals thatsurround us and, we must assume, the very act of suitably naming them played a criti-cal role in our linguistic and cultural evolution.
We continue to recognize the emotive power of linking creature and name-of-creature in the zoological lexicon of current-day local languages spoken by members
of traditional societies. It is likely, as Köhler and others have suggested, that markingnatural kinds with verbal labels that reflect their non-arbitrary inherent properties ismore apparent in the languages of small, local populations than in the modern worldlanguages of technologically complex populations. And it should hardly be news thatthe focus on form (shape), movement, and relative size is so apparent, given that thesethree perceptual dimensions mark the most significant inherent properties of livingthings that humans immediately recognize and aim to represent symbolically in a crea-ture’s name (see Adams & Conklin ; Clark ).
From the conservative perspective of historical linguistics, it is somewhat discon-certing that phonaesthesia remains so strong a nomenclatural principle in spite of thelevelling effects of regular sound change and the subsequent loss of sound-meaningassociations that accompanied the development of duality of patterning – perhaps oneof the most important signposts in language evolution. But good, workable solutionsare not easily abandoned – as pamau and wiching for tapir and squirrel, tungkau andíspik for giant doradid catfish and little angelfish, and wa and kawachaa for tinamouand rail so clearly demonstrate. These are as good names as we are likely to find in thenaming game and, should they some day be replaced, those that follow will be equally intuitively satisfying, to both our eyes and ears. Thus, by carrying on in the traditionof our colleagues at the First Congress of Ethnozoological Nomenclature, our namesfor animals continue to mirror the life of the world of nature to which they have been,and will continue to be, symbolically assigned.
NOTES
This paper owes much to the critical comments of John Ohala. His proposals on the ‘frequency code’ underlying sound symbolism (Ohala ) forms the empirical foundation for all serious research onthe topic. I have not always taken his advice and any major errors on acoustic phonetics are my own.Allen Jensen and other researchers of the Summer Institute of Linguistics have provided important origi-nal data from their long-term linguistic work with South American Indian languages. This paper could not have been written without access to their resources. I also wish to note my appreciation for the supportand encouragement of Roy Ellen. I am grateful to him for his kind invitation to read an earlier version ofthis paper in his symposium at the Ninth International Congress of Ethnobiology and for his cogent edi-torial advice in helping me move a rough oral presentation into its current written form. Finally, I thankElois Ann Berlin, whose useful and ingenious suggestions helped me design the experiments described inthe text.
1 Since the times of Plato’s Cratylus ( BC), it has been recognized that names in natural languagesshould in some way be non-arbitrarily assigned to their referents.
Then [says Socrates to Hermogenes], I should say that this giving of names can be no such lightmatter as you fancy … and Cratylus is right in saying that things have names by nature, and that notevery man is an artificer of names, but he only who looks to the name which each thing by naturehas, and is able to express the true forms of things in letters and syllables (Plato n.d.).
This position contrasts markedly with the prevailing view in modern linguistics that flows directly from Fer-dinand de Saussure’s doctrine on the arbitrary relationship of speech sounds and meaning. The strength ofde Saussure’s views has been so pervasive that serious studies of sound symbolism were relegated to thebackwaters of linguistic theory until recently (Ciccotosto ; Foster ; Hinton, Nichols & Ohala ;Linguistic Iconism Association n.d.; Nuckolls ; ). The empirical research available on the non-arbitrariness of the relation of sound and meaning in language now demonstrates that the standard doctrine is surely inadequate. One is no longer accused of practising fringe linguistics by conducting researchon sound symbolism as a pervasive semantic process in language.
2 My paper is dedicated to the late social psychologist Roger Brown, who published an important paperwith this title (b). His work on the psychology of language, and especially sound symbolism, has con-tinued to inspire me since the time I first read his Words and things (a).
3 Páipaich is the name of Lipaugus vociferans in Aguaruna, a Jívaroan language of Amazonas, Peru.4 It is appropriate to recognize here the work of Dwight Bolinger and Yakov Malkiel, primary proponents
of sound-symbolic studies in mid-twentieth-century American structural linguistics. Bolinger coined theterm ‘phonosemantics’ (), whereas Malkiel, focusing primarily on historical problems, preferred theterm ‘phonosymbolism’ (). Bolinger credits Tolman (, ) for having conducted sound symbol-ism research prior to that of Sapir (Bolinger, personal communication).
5 This high-front unrounded vowel is written phonetically [i], as pronounced in English see, Spanish mil‘one thousand’, or Japanese mi ‘fruit, nut’.
6 In his honour, Professor Uznadze is remembered today by the Dimitry Uznadze Institute of Psychology,Tbilisi, Republic of Georgia.
7 So far as I have been able to discover from a review of the extensive experimental literature using theKöhler figures – and it is vast – this is the first experiment that required subjects to produce their own namesfor the drawings.
8 I appreciate the collaboration of Richard Shedenhelm, Department of Philosophy, for allowing me touse the students in his classes for this experiment.
9 Note that the three major perceptual dimensions that focus on movement, shape, and relative size areprecisely the basic organizational properties of objects that humans acquire in their categorization of theworld in general (see Adams & Conklin ; Clark ).
10 After producing this summary table, I discovered in a citation by Fox (: ) that the well-knownAfricanist anthropologist M. Westermann () had published a remarkably similar table.
11 I thank Professors Mikel Gleason and Mark Williams for allowing me to use students in their intro-ductory classes for this experiment. For the present paper, I report only on the results from the experimentconducted in Professor Gleason’s class.
12 My specific instructions were, ‘After looking closely at the birds, under each bird write down a Droidesename that you believe best represents their inherent qualities, names that just seem natural to you’.
13 Paget’s ideas on vocal gestures were apparently inspired by his work with deaf children and his efforts(with Grace Paget and Pierre Gorman) to develop a manual sign language, referred to today as Paget-Gorman Signed Speech.
REFERENCES
A, K.L. & N.F. C . Toward a theory of natural classification. In Papers from the NinthRegional Meeting of the Chicago Linguistic Society (eds) C. Corum, T.C. Smith-Clark & A. Wieser, -.
B, B. . Ethnobiological classification: principles of categorization of plants and animals in traditionalsocieties. Princeton: University Press.
——— . Evidence for pervasive synesthetic sound symbolism in ethnozoological nomenclature. InSound symbolism (eds) L. Hinton, J. Nichols & J. Ohala, -. Cambridge: University Press.
——— . Tapir and squirrel: further nomenclatural meanderings toward a universal sound-symbolicbestiary. In Nature knowledge: ethnoscience, cognition, and utility (eds) G. Sanga & G. Ortalli, -. NewYork: Berghahn Books.
——— . ‘Just another fish story?’ Size-symbolic properties of fish names. In Animal names (eds) A.Minelli, G. Ortalli & G. Sanga, -. Venice: Istituto Veneto di Scienze, Lettere ed Arti.
B, D. . The sign is not arbitrary. Boletin del Instituto Caruso y Caro , -.B, R. a. Words and things. New York: Free Press.——— b. How shall a thing be called? Psychological Review , -.———, A.H. B & A.E. H . Phonetic symbolism in natural languages. Journal of Abnormal
and Social Psychology , -.——— & R. N . Method in phonetic symbolism experiments. Journal of Abnormal and Social
Psychology , -.B U n.d. Website (http://www.butterflyutopia.com/.html).C, L. . The annotated Alice: Alice’s adventures in wonderland and through the looking glass (with
introduction and notes by M. Gardner). New York: Charkson N. Potter, Inc.C, N. . Sound symbolism in natural language. Ph.D. dissertation, University of Florida,
Gainesville, Florida.C, E.V. . Classifiers and semantic acquisition: universal categories? Paper presented at the rd
annual meeting of the American Anthropological Association, Mexico City, - November.D, R. . The fitness of names to drawings: a cross-cultural study in Tanganyika. British Journal of
D U n.d. Website (http://www.slaxicon.org/ulwa/dict/).D, M. . Origins of the modern mind: three states in the evolution of culture and cognition.
Cambridge, Mass.: Harvard University Press.——— . Précis of Origins of the modern mind: three stages in the evolution of culture and cognition. Behav-
ioral and Brain Sciences , -.——— . A mind so rare: the evolution of human consciousness. New York: W.W. Norton and Company.F, G. . Speech sounds and features. Boston: MIT Press.F, G., R. W & D. F . Diccionario Achuar-Shiwiar-Castellano. (Serie Lingüística Peruana
). Lima: Ministerio de Educación/Instituto Lingüístico de Verano.F, M.L. . Body processes in the evolution of language. In Giving the body its due (ed) M. Sheets-
Johnstone, -. Albany: State University of New York Press.F, C.W. . An experimental study in naming. American Journal of Psychology , -.H, H. . Diccionario Chayahuita-Castellano (Serie Lingüística Peruana ). Lima: Ministerio de
Educación/Instituto Lingüístico de Verano.H, S.L. . Birds of Venezuala. (Second edition). Princeton: University Press.H, L., J. N & J. O (eds) . Sound symbolism. Cambridge: University Press.H, M. . The interplay of metaphor and iconicity: a cognitive approach. Unpublished Ph.D. thesis,
University of London.I, D. . Resource management by the Runa Indians of the Ecuadorian Amazon. Ph.D. dissertation,
Stanford University, Stanford, California.I, F.W. & E. N . A genetic study of the naming of visual figures. Journal of Psychology ,
-.J, R. . Six lectures on sound and meaning. Cambridge, Mass.: MIT Press.——— & R. W . The sound shape of language. Berlin: Mouton de Gruyter.J, A.A. . Sistemas indígenas de classificação de aves: aspectos comparatives, ecológicos, e evolutivos.
Belém-Pará: Ministério de Cîencia e Tecnologia, Conselho Nacional de Desenvolvimento Científico e Tecnológico, Meseu Paraense Emílio Goeldi.
J, O. []. Symbolic value of the vowel i. In Linguistica, -. College Park, Md.K, W. . Gestalt psychology. New York: Liveright Co.——— . Gestalt psychology. (Second edition). New York: Liveright Co.L, P. . Vowels and consonants: an introduction to the sounds of languages. Oxford: Blackwell.——— n.d. A course in phonetics (available on-line: http:hctv.humnet.ucla.edu/departments/linguistics/
VowelsandConsonants/course/chapter/.htm).L, G. & M. J . Metaphors we live by. New York: Barnes and Noble.L, G., Z. L, J. B & J. L B . Messa di voce: an audiovisual performance and instal-
lation for voice and interactive media (available online: http://tmema.org/messa/messa.html#background).L, M.S. . Size and distance perception of the physiognomic stimulus taketa. Bulletin of the
Psychonomic Society : , -.——— . The meanings of the physiognomic stimuli taketa and maluma. Bulletin of the Psychonomic
Society : , -.Linguistic Iconism Association n.d. Website (http://www.conknet.com/~mmagnus/LIA/).L, E. & B. L . Diccionario Capanahua-Castellano (Serie Lingüística Peruana ). Lima: Ministe-
rio de Educación/Instituto Lingüístico de Verano.MN, P. . Prolegomena to a theory of the sound pattern of the first spoken language. Phonet-
ica , -.——— . The frame/content theory of evolution and speech production. Behavioral and Brain Sciences
, -.——— & B.L. D . On the origin of internal structure of word forms. Science , -.——— & ———. Motor mechanisms in speech ontogeny: phylogenetic, neurobiological and linguis-
tic implications. Current Biology , -.M, Y. . Diachronic problems in phonosymbolism: edita and inedita, -, vol. . London: John
Benjamins.M S, R. & W.H. P . A guide to the birds of Venezuela. Princeton: University
Press.M, S. . The prehistory of the mind: a search for origins of art, religion and science. London: Thames
& Hudson.N, J. . Diminutive consonant symbolism in western North America. Language , -.
N, J.B. . Sounds like life: sound-symbolic grammar, performance, and cognition in PastazaQuechua. Oxford: University Press.
——— . The case for sound symbolism. Annual Review of Anthropology , -.O, J. . An ethnological perspective on common cross-language utilization of F0 of voice. Phonet-
ica , -.——— . The frequency code underlies the sound-symbolic use of voice pitch. In Sound symbolism (eds)
L. Hinton, J. Nichols & J. Ohala, -. Cambridge: University Press.O, F. & F. Q n.d. Ornitología Cuiva-Guahibo (available on-line: http://www.vjf.cnrs.fr/celia/
FichExt/Am/A__.htm).P, R. . Human speech: some observations, experiments, and conclusions as to the nature, origin,
purpose, and possible improvement of human speech. London: Routledge & Kegan Paul.P n.d. Cratylus (trans. B. Jowett; available on-line: http://eserver.org/philosophy/plato/cratylus.txt).R, V.S. & E.M. H . Synaesthesia: A window into perception, thought, and lan-
guage (available on-line: http://psy.ucsd.edu/chip/pdf/Synaesthesia%-%JCS.pdf ).R, R. . Diccionario Arabela. (Serie Lingüística Peruana ). Lima: Ministerio de Educación/Instituto
Lingüístico de Verano.SAILDP (South American Indian Languages Documentation Project). n.d. Data files available in personal
data files of B. Berlin, Laboratories of Ethnobiology, University of Georgia, Athens, Georgia.S, E. . Diminutive and augmentative consonant symbolism in Wishram. In Handbook of American
Indian languages, Bull. , Part I, -. Washington, D.C.: Bureau of American Ethnography.——— . A study in phonetic symbolism. Journal of Experimental Psychology , -.S, M. . On inter-hemispheric linguistic connections. In Culture history: essays in honor of Paul
Radin (ed) S. Diamond, -. New York: Columbia University Press.T, W. & E. T . Diccionario Bora-Castellano-Castellano-Bora. (Serie Lingüística Peruana
). Lima: Ministerio de Educación/Instituto Lingüístico de Verano.T, A.H. . The laws of tone color in the English language. Andover Review , -.——— . Symbolic value of English sounds. In The views about Hamlet, and other essays, A.H. Tolman.
New York: Houghton Mifflin.T, R. . What makes sound patterns expressive? The poetic mode of speech-perception. Durham, N.C.:
Duke University Press.——— . Onomatopoeia: cuckoo-language and tick-tocking: the constraints of semiotic systems (avail-
Size-Sound_Symbolism.pdf ).T, J. . Vocabulario Candoshi de Loreto. (Serie Lingüística Peruana ). Lima: Ministerio de
Educación/Instituto Lingüístico de Verano.U, R. . Size-sound symbolism. In Universals in human language, vol. : Phonology (ed.) J.
Greenberg, -. Stanford: University Press.U, D. . Ein experimentaller Beitrag zum Problem der psychologischen Grundlagen der Namen-
gebung. Psychologische Forschung , -.V, J. []. Language: a linguistic introduction to history (trans. P. Radin). New York: Barnes &
Noble.V, A. n.d. Dicionário Jarawara-Português. Sociedade Internacional de Lingüística, Cuiabá, MT (avail-
able on-line: http://www.sil.org/americas/brasil/PUBLCNS/DICTGRAM/JADict.pdf).W, D. . Laut, Ton und Sinn in westafrikanische Sudansprachen. In Festschrift Meinhof, -
. Hamburg: Kommissionsverlag von L. Friederichsen & Co.W, A. . Gantëya bain: el pueblo Siona del río Putumayo, Colombia, vol. II. Lomalinda, Meta
Colombia: Editorial Townsend.
Pour un premier Congrès de Nomenclature ethnozoologique
Résumé
Les dénominations indigènes désignant les principales catégories ethnobiologiques s’avèrent régies pardes principes de nomenclature réguliers. Il est bien établi que deux de ces principes à l’œuvre dans lanomenclature ethnozoologique, l’onomatopée et la description métaphorique de telle ou telle propriété
observable de l’organisme désigné, constituent la base des désignations génériques vernaculaires. Oncomprend moins bien un troisième principe, que l’on a pu appeler « symbolisme sonore ». Dans leslangues des peuples traditionnels, les noms d’animaux dont la signification reste obscure ont souvent despropriétés de symbolisme sonore, associées inconsciemment à l’essence ou à la nature de la créature enquestion. Comment expliquer ceci après tous les changements subis par la communication verbalehumaine depuis la naissance de ce que l’on pourrait appeler le langage à part entière ? De quelle manièreces principes sont-ils liés aux principes généraux de classification naturelle basés sur la forme et le mou-vement ? Si l’imitation verbale constitue un stade critique dans l’évolution de la cognition humaine,quelles spéculations éclairées pourraient en émaner qui viendraient alimenter les débats d’un premierCongrès de Nomenclature ethnozoologique ?
Brent Berlin is Professor of Anthropology at the Laboratories of Ethnobiology, Department of Anthropol-ogy, University of Georgia, USA.
Department of Anthropology, University of Georgia, Athens, GA , USA. [email protected]
