1
2
3	IDEOPHONES WORLDWIDE It has long been known that a feature characteristic of languages worldwide, but particularly those of Africa, is ideophones, words of a distinct semantic type, which may fill one or many syntactic slots. Ideophones may be defined as a subset of sound symbolism, which also includes phonaesthemes and other methods of indicating qualities (for example alternations of ± ATR vowels) This field is often referred to as phonosematics and has a long history in Western philosophy. Plato’s Cratylus has a discussion of phonaesthemes, for example Ideophones (or ‘expressives’ in Asian terminology) have begun to be of more interest to the broader scholarly community (e.g. Hinton et al, 1994).
4	IDEOPHONES WORLDWIDE Ideophones are abundant in natural and heightened speech, notably in Africa, but absent from typical example sentences, hence their failure to be treated adequately in typical grammars and dictionaries. They are hard to elicit since their existence is unpredictable and speakers have no natural ‘hook’ to recall them Their elusive nature, in grammatical terms, has made them poor relations to other word classes and they have been little treated by the schools of grammar dominated by syntax (see Welmers 1973 for comment on lacunae in research).
5	IDEOPHONES in English Typical examples in English are; Shilly-shally (verb) Hocus-pocus (noun) Namby-pamby (adjective) Dingdong (onomatopoeia) Helter-skelter (adverb/noun)
6	IDEOPHONES WORLDWIDE The first clear reference to a class of ideophones was in 1886, where they are called ‘indeclinable verbal particles’ (McLaren 1886). Banfield (1915) whose documentation for Nupe is particularly rich, calls them ‘intensitive adverbs’. Doke (1935) called them ‘a vivid representation of an idea in sound’. Detailed studies such as Kunene (1978) on Southern Sotho suggest that some Niger-Congo languages may have thousands of such ideophones Our understanding of the role they play in natural language (as opposed to elicited examples) is still very preliminary.
7	IDEOPHONES WORLDWIDE It seems that ideophones are far more prevalent in Africa than in other world language phyla. It is hard to be sure, but lists of ideophones for other predominantly oral regions of the world seem to be remarkably short Possibly because other strategies (such as a rich repertoire of adjectives or adverbs or phonoaesthemes) substitute. However, language phyla also just differ and Africa may be a special case
8	IDEOPHONES WORLDWIDE The classification of ideophones remains under debate. They have been defined very broadly in the literature as anything with a sound-symbolic element, in which case they can be found in all major parts of speech. For example, English verbs such as ‘gobble’ or ‘twinkle’ are sometimes treated as ideophones In this view, being ideophonic is more a matter of conforming to a certain syllable structure. However, in many languages, specific words, which may or may not be morphologically marked, fill a slot that would usually be adverbial.
9	IDEOPHONES WORLDWIDE In many languages, ideophones have distinctive phonotactics, although this is not always the case. But they do always have highly specific applications to the sensory world and describe visual, aural and emotional experiences in ways hardly paralleled elsewhere in the lexicon. Historically, they are hard to treat, as they do not seem to be lexically cognate across languages. There is one intriguing exception to this, the worldwide word for ‘round, circle, wheel’ which is often k-l- or k-r- in many language phyla.
10	IDEOPHONES WORLDWIDE In contrast, semantically, sensory experiences can be identified across languages and even phyla. Many African languages have ideophones describing comparable experiences, for example, the different noises made by objects falling on the gorund. If so, then ideophones are crucial to a broader understanding of the perceptual world implicit in African languages.
11	IDEOPHONES references Samarin (1965, 1967, 1971) is one of the few authors to give Niger-Congo ideophones a thorough discussion. He has surveyed Bantu ideophones but has also explored the wealth of ideophones in Gbaya, a Ubangian language. Other discussions can be found in Evans-Pritchard (1962) for Zande, Courtenay (1976) for Yoruba, Fivaz (1963) and Von Staden (1977) for Zulu, Fortune (1962) for Shona, Nurse (1974) for Nyanja, Geunier (1978) for Malagasy, Hulstaert (1962) for Mongo, Madugu (1987) for Nupe, Mamet (1978) for Ntomba, Noss (1975, 1986) for Gbaya, Uzechukwu (1980) for Igbo, Marivate (1985) for southern Bantu, Mphande (1989), for Tumbuka, Kulumeka (1993, 1997) for Chewa.
12	IDEOPHONES in africa A great deal less is known about Nilo-Saharan and Afroasiatic although Chadic languages are clearly as rich as their Niger-Congo neighbours Although it is not entirely clear, dictionaries suggest that Afroasiatic and Khoesan are less replete with ideophones and that it is interaction with Nilo-Saharan and Niger-Congo that increases their repertoire This is curious since there is no evidence of direct lexical borrowing; therefore it is essentially borrowing of ideas about expressivity
13	IDEOPHONES in africa An aspect of ideophones in most African languages is reduplication; words are wholly or partly reduplicated according to language-internal rules and it is often these reduplications that give the onomatopoeic sense that plays some role in their generation. Ideophones tend to be polysyllabic and in some cases can be freely extended. However, the amount of reduplication varies from language to language; in Kanuri, for example, where reduplication is not a very common process in the language as a whole, many ideophones are not of this form There is probably a very general relationship between canonical structures of words in a language and the form of ideophones
14	Claims about ideophones Ideophones are not usually amenable to historical linguistics; their etymologies are generally opaque and their is little cross-language regularity. One intriguing question revolves around renewability and innovation. Do speakers constantly invent new ideophones to suit changing environments? Does the pool of ideophones constantly renew itself over time faster than the replacement rate of ordinary lexemes? Speakers generally claim that ideophones are a fixed pool and cannot be just ‘made up’. The exception is imitations of sounds; for example new technologies require new ideophones
15	Claims about ideophones A common claim in the literature on ideophones in African languages is that these are phonologically marked. They can contain unusual consonant phonemes (less usually vowels) or sequences. They also more often contain glide tones than the ordinary lexicon The labio-dental flap, recently recognised by the IPA, is more common in many languages in ideophones than in ordinary words. Indeed, its recognition depended on the argument that it is common in ‘ordinary’ words in Mambay Courtenay (1976) argues that phonological markedness is the case in Yoruba, as does Madugu’s (1987) for Nupe. This is certainly the case for many of the southern African Bantu languages studied in detail, where specific rules of reduplication and tone-patterns abound.
16	Claims about ideophones A claim frequently made in older literature is that ideophones can be constructed by individual speakers to suit a particular speech-event and have no language-wide validity. As far as this can be tested, there appears to be no validity for this in the case of Tarok. Ideophones as defined here have a particular syntactic slot and cannot be easily invented and understood. This is not to deny that individuals cannot imitate, sometimes remarkably effectively, new auditory experiences. However, these are regarded as outside the language system proper.
17	Some claims about Tarok ideophones 1. They cannot be distinguished from adverbs morphologically or syntactically 2. are not morphologically distinct from the main Tarok lexicon 3. have no unusual phonological properties 4. are not tonally distinct from similar non-ideophones 5. have etymologies that can only very rarely be discerned 6. constitute a fixed set of forms known to all competent speakers 7. Therefore, if they are to be distinguished as a word-type it is only through semantics
18	The interface with other types of ‘sensory’ phones I 1. Many African languages also have ophresaesthemes, words to describe very specific smells. A popular one in Nigeria/Cameroun is the ‘smell of fresh dogmeat’, admittedly not common in European contexts. 2. These do not fill the same syntactic slot as ideophones and behave more like invariant nouns. Nonetheless they appear to fill the same experiential slot as ideophones.
19	The interface with other types of ‘sensory’ phones II 1. Across Central Nigeria, many Plateau and neighbouring languages have what may be called ‘insultatives’ 2. These are invariant adjectives that qualify particularly body parts and are only used in insults. 3. They do not resemble ideophones morphologically in languages where this is marked and do not show concord in languages where other adjectives do. 4. Nonetheless, they otherwise appear to fall into the same experiential class as ophesaesthemes etc.
20	Conclusions Our knowledge of the extent of ideophones in particular African languages is in part because their recording reflects the worldview of the compiler of the lexicon. Our knowledge of their use is often highly defective even for language where the lexical forms have been documented because of the way we write grammars. It is clear that ideophones can be phonotactically, morphologically or syntactically marked, but this is not a necessary requirement It is probably better to treat them as an ‘experiential’ class, some thing which describes and intensifies the interaction with the sensory world and to acknowledge that African languages at least also have other related word classes which also remain in descriptive limbo