Early sources on African ideophones, part I: Schlegel on Ewe, 1857

This is the first post in a series. Featured philologist of today is Joh. Bernhard Schlegel, for providing us with precious data on ideophones (expressives) in nineteenth-century Ewe, a Kwa language of southeastern Ghana. But since this is the first post on ideophones here, let’s first try to answer the obvious question: what are ideophones, anyway?

Ideophones 1 are a type of words used by speakers to convey a vivid impression of a certain sensation or sensory perception. They are found abundantly in Asian and African languages, as well as in some South American languages. It is important to note that in these languages, they form a distinct class of words, definable by a constellation of phonological, morphological, syntactic and semantic criteria. As a class of words, they are rare in Indo-European languages.

Cross-linguistically, ideophones are marked words in many ways. 2 They are phonologically marked by deviant phonotactics and all sorts of co-occurrence constraints; they have special morphology (often iconic, e.g. reduplication, lengthening); there is some syntactical ‘aloofness’ to them (can’t be negated, are often less well integrated into the sentence); and they are semantically very specific, typically evoking experiences as a whole rather than encoding some attributes of objects of events. Part of what makes them stand out is that ideophones utilize sound symbolism to map onto the ‘analogue’ sensory world.

Now that we have at least a vague idea of what ideophones are, let’s move on to Schlegel. Joh. Bernhard Schlegel was one of the first missionaries of the Norddeutsche Missionsgesellschaft to arrive in what is now southeastern Ghana (around 1847). He lived and preached among the Ewe and in the process managed to amass quite an amount of ethnographic and linguistic information. The linguistic part was published 150 years ago in the form of the aptly-titled Schlüssel der Ewesprache. 3 Schlegel’s Schlüssel is a combined grammar and dictionary of the Anlo dialect of Ewe, a Gbe language of the Niger-Congo phylum, Africa’s biggest language family. What does it have to tell us about ideophones?

Schlüssel der Ewesprache, §15: Adverbien

Chapter 15 (pp. 108-117) treats a wide range of words under the heading ‘Adverbien’. Most interesting for us is section 4 on Adverbien der Intensität und der Frequenz. Indeed, in section 4 Schlegel writes the following:

Eine besondere Art von Intensitäts- und Frequenzadverbien, welche eine sehr reiche Entfaltung erfahren hat, ist zu erwähnen, Adverbien, welche zur näheren Bestimmung eines bestimmte Adjektivs oder einer bestimmten Begriffsgruppe verwendet werden. Die Sprache hat beinahe zu jeglichem Eigenschafts- oder Beschaffenheitsbegriff ein eigenes derartiges Adverb gebildet, welches je nach dem Grade der Intensität, den es bezeichnen soll, redupliziert wird. (Schlegel 1857:113)

A special type of adverbs of intensity and frequency, which has seen a very rich development (lit. ‘unfolding’), is to be noted, namely adverbs that are used to further define a specific adjective or a specific group of concepts. For almost every property or quality-concept, the language has developed such an adverb, which is reduplicated according to the degree of intensity to be conveyed.

In short: in Ewe, there is a special type of adverb; there are many of them; they serve to modify property concepts (‘Beschaffenheitsbegriff’), which apparently includes both stative verbs and adjectives; and they can be reduplicated to reflect the degree of intensity. What do these words look like? Schlegel gives quite a few nice examples, of which I’ll cite just four. (I adjust his transcription to modern-day Ewe orthography, though without correcting transcription errors (e.g. d for /ɖ/) and without adding tone marks.) 4

  1. eŋũti kɔ̀ kaŋkaŋkaŋ ‘es glänzt golden, schimmernd u. s. w.’ [it shines golden, glimmering, etc.]
  2. eʋó ʋɔ̀ kó-kó-kó ‘er klopfte sachte an’ [he knocked at the door gently]
  3. eʋó ʋɔ̀ gbɔ̀gbɔ̀gbɔ̀gbɔ̀ ‘er klopfte heftig, dreist, beharrlich an’ [he knocked at the door boldly, firmly]
  4. eʋè lilililili, ‘es riecht fein, angenehm’ [it smells fine, pleasantly]

One thing to note here is the sound-symbolic opposition of voiceless and voiced consonants in examples 2 and 3, reflecting intensity of sound. This is a very common phenomenon in ideophones. Example 4 is also interesting because we also have another variant of the smell ideophone: lìlìlìlì ‘smelling bad’ (Westermann 1927, Ameka 2001). That is, a tonal contrast High-Low maps onto an evaluative contrast good-bad; another instance of sound-symbolism (see Westermann 1927 for more examples). Schlegel doesn’t actually give the low-tone variant; this may be just a gap in his data.

Whence these words?

In the light of Schlegel’s words above concerning the sehr reiche Entfaltung of this type of words, one might ask: whence these words, and why so many of them? Schlegel has something to say on that, too:

Wie bereits bemerkt, so hat die Eʋe-Sprache einen fast unerschöpflichen Reichthum an Adverbien dieser Art entfaltet. Der Grund dieser Erscheinung liegt darin: einmal, daß der Eʋeer alle denkbaren Modifikationen der Begriffe anschaulich zu machen pflegt, dem Denken muß auch die Aktion in der äußeren Erscheinung und den Geberden entsprechen, welcher hinwiederum die Sprache in bezeichnenden und entsprechenden Ausdrücken folgt … (Schlegel 1857:114)

As was already noted, the Ewe language has developed an almost inexhaustible abundance of this type of adverbs. The basis for this phenomenon lies herein: first, [in the fact] that the Ewe wants to illustrate all conceivable modifications of concepts; the thinking is followed by actions in the form of outer appearances and gestures, which in turn are accompanied by language in the form of signifying expressions accordingly 5

Thus, ideophones arise from an urge to illustrate alle denkbaren Modifikationen der Begriffe. Importantly, ideophones are not just words; they come with actions in the form of outward appearances and gestures — in short, they are multi-modal. What emerges here from Schlegel’s rather dense prose is some sort of a miniature model of gesture and speech production. It is not all that sophisticated of course, but all the same it is quite amazing to see what careful observation could yield one and a half century ago. Schlegel also observes that the gestures here are time-aligned with the speech they accompany. This is not noted very often in connection to ideophones, especially not in older (pre-1950) sources; but I think we can safely attribute this to the common failure of linguists to look beyond the spoken word rather than to absence in the actual data.

So Schlegel is way ahead of his time here 6; the role played by gestures in expressive speech has never been widely recognized, and although some authors have hinted at its importance (e.g. Madan 1911, Samarin 1971:153, Diffloth 1972:144), serious studies of the tight coupling between ideophones and spontaneous iconic gestures are not older than ten years (most importantly, Kita 1997). With this in mind, I can only say (using one of these strange words, and leaving the room with a bowing gesture): akpe kakakakaka — thank you, Joh. Bernhard Schlegel!


  1. Ameka, Felix K. 2001. Ideophones and the Nature of the Adjective Word Class in Ewe. In Ideophones, ed. F. K. Erhard Voeltz and Christa Kilian-Hatz, 25-48. Amsterdam: John Benjamins.
  2. Diffloth, Gérard. 1972. Notes on expressive meaning. Chicago Linguistic Society 8:440–447.
  3. Kita, Sotaro. 1997. Two-dimensional semantic analysis of Japanese mimetics. Linguistics 35:379-415.
  4. Madan, A C. 1911. Living Speech in Central and South African: an Essay Introductory to the Study of the Bantu Family of Languages. Oxford: Clarendon Press.
  5. Samarin, W. J. 1971. Survey of Bantu ideophones. African Language Studies 12:130-168.
  6. Schlegel, Joh. Bernhard. 1857. Schlüssel der Ewesprache, dargeboten in den Grammatischen Grundzügen des Anlodialekts. Stuttgart.
  7. Schlegel, Joh. Bernhard. 1858. Beitrag zur Geschichte, Welt- und Religionsanschauung des Westafrikaners, namentlich des Eweer. Monatsblatt der Norddeutsche Missionsgesellschaft 7:398ff.
  8. Westermann, Diedrich Hermann. 1927. Laut, Ton und Sinn in Westafrikanischen Sudansprachen. In Festschrift Meinhof, 315-328, Hamburg: J.J. Augustin.


  1. Expressives and mimetics are two other commonly used terms for the same phenomenon.[]
  2. Many of the ideas in this brief overview were articulated in close collaboration with my colleague Sylvia Tufvesson, who is working on expressives in Semai, a Mon-Khmer language of the Malay peninsula.[]
  3. Schlegel actually writes Ew̉esprache, marking the voiced bilabial fricative /ʋ/ with a hook above the w (see also note 3 below). However, it is always cited without that diacritic and I will conform to that tradition here.[]
  4. Most important changes: w̉ → ʋ, n ̇ → ŋ, dš → dz, tš → ts, e → ɛ, o → ɔ []
  5. The German here is really difficult to parse, even for native speakers (which I am not); but apparently the suggestion is that this is a back-and-forth between though, action, and language.[]
  6. Of course, the picture is not that idyllic. Schlegel’s analysis may be ahead of his time, yet he shows himself a true child of his age when in the same breath he closes off the section with the following words: ‘… dann aber auch darin, daß die Entfaltung der Adjektiven und der eigentlichen Adverbien noch auf einer anfänglichen Stufe zurück ist.’ (pp. 114-5, emphasis mine; translation: “…but also in the fact that the development of the adjectives and the real adverbs still finds itself in an initial stage.”). For those readers who were growing afraid that Ewe actually might be a highly developed language: it’s still just a backward stage of development, baby.[]

9 thoughts on “Early sources on African ideophones, part I: Schlegel on Ewe, 1857”

  1. Great post. I realize this makes for a lot more work for you, but a humble request from a non-Germanophone reader: might you, in the future, be able to provide English translations for your German quotes?

  2. I got another request for translation by email, so I think I’ll be translating any German quotes in the future. A translation of the ones in this post will follow soon. Thanks for the suggestion!

  3. Hello Mark,
    This is great, although there are some things in here that call for discussion and more precision. Let me mention one. I am sure you are aware of it. I think expressives or ideophones are a LEXICAL class but they are not a word class. I think languages vary in this. For some languages, Expressives or Ideophones may belong to only one word class; for others they are divided over words and still in others they are a subclass of one or more classes.
    I wonder why some of your special symbols do not show up on my machine. Some show without problem it is really in the notes that the problem shows up.
    Season’s greetings

  4. Hi Felix,

    Yes, point taken about the lexical class / word class distinction. I agree that a language can have expressive (ideophonic) vocabulary without having a word class (in the grammatical sense) of expressives. In fact, I’ve been working on a post about expressive verbs in Tuareg, so Tuareg would be an example of a language of your third type (in which expressives occur as a lexical subclass within a word class).

    As for the special symbols, I think they should display alright now even in the footnotes. I had forgotten to apply a special layout to them, thanks for pointing that out!

  5. Schlegel also observes that the gestures here are time-aligned with the speech they accompany

    This is interesting, but not restricted to ideophones; I’ve noticed that English gestures tend to start and end along with intonation groups, and “peak” at stressed syllables.

    expressive verbs in Tuareg

    I’ve just lately been looking at something rather similar sounding in Kwarandjie, so I’m curious to hear more…

  6. This is interesting, but not restricted to ideophones; I’ve noticed that English gestures tend to start and end along with intonation groups, and “peak” at stressed syllables.

    You are quite right, and this has been noted for gesture in general (McNeill 1992). What makes this piece of information interesting is that on the whole, expressives tend to come together with gestures more often than other words (e.g. ‘normal’ nouns or verbs).

    I’ll let you know when I post the bit about expressive verbs in Tuareg. I’m looking forward to your Kwarandjie data!

Leave a Reply

Your email address will not be published. Required fields are marked *