Abstract
This paper examines the use of ideophones in motion event descriptions in Kupsapiiny (Southern Nilotic; Uganda; anti-verb-framed with path-affix satellites) and Sidaama (Highland East Cushitic; Ethiopia; verb-framed). Both languages lack verbs for expressive manners of motion, but Kupsapiiny has a much richer inventory of expressive ideophones that can be used to describe motion events than Sidaama. This contradicts Schaeferâs (2001) hypotheses that ideophones are largely restricted to verb-framed languages and that they primarily depict expressive manners of motion, which are absent from these languagesâ lexical inventories. Although ideophones might seem irrelevant to Talmyâs (2000) motion typology, a closer examination reveals that while both languages lack direct means of expressing detailed manners through verbs, adverbs, or ideophones, Kupsapiiny distinguishes various sounds using ideophones, only some of which are loosely associated with specific motions, whereas Sidaama often conceptualizes unusual manners as postures, with ideophones only occasionally used to describe posture changes.
1 Introduction
This study investigates the use of ideophones in motion event descriptions in two African languages, Kupsapiiny (Kalenjin, Southern Nilotic; Uganda; iso 639-3: kpz) and Sidaama (Highland East Cushitic; Ethiopia; iso 639-3: sid), which differ substantially in Talmyâs (1985, 1991, 2000) motion typology. Kupsapiiny is predominantly anti-verb-framed with path affix satellites, while Sidaama is verb-framed. In both languages, ideophones appear in quotative constructions but differ considerably. One striking difference is that Kupsapiiny ideophones are âprimaryâ (Dwyer & Moshi, 2003) and âexpressiveâ (Dingemanse & Akita, 2016; Dingemanse, 2017), most of which represent sound, whereas Sidaama ideophones are conventionalized or idiomatized and somewhat âgrammaticalizedâ (Dwyer & Moshi, 2003; Andrason & Heine, 2023) or âgrammatically integratedâ (Dingemanse & Akita, 2016; Dingemanse, 2017), without restriction to any particular semantic domain.
This study first examines Schaeferâs (2001) hypotheses (i) that the use of ideophones for manners of motion is restricted to verb-framed languages and (ii) that in verb-framed languages, ideophones convey expressive manners of motion, for which Slobin (1997, 2000, 2004) argues such languages lack verbs in their lexical inventories.1 The findings show that, contrary to these hypotheses, like Sidaama, Kupsapiiny also lacks verbs for expressive manners of motion; however, Kupsapiiny has a much larger inventory of expressive ideophones that can be used in motion event descriptions than Sidaama. This suggests that Schaeferâs hypotheses are untenable.
The study also examines video description data and shows that both languages use ideophones only infrequently for motion events with ordinary manners, but that Kupsapiiny frequently employs ideophones for motion events with unusual manners, whereas Sidaama does not. A closer analysis indicates that, although neither language has direct means to express detailed manners through verbs, adverbs, or ideophones, speakers conceptualize motion events with unusual manners not necessarily as events with manners, but rather often as events involving other types of event components: Kupsapiiny distinguishes various sounds using ideophones, only some of which are loosely associated with specific motions, while Sidaama often conceptualizes unusual manners as posture changes, with ideophones only occasionally used to describe these posture changes. In Talmyâs motion typology (1985, 1991, 2000), the figure entityâs production of sounds or gradual posture change during motion can be analyzed as an instance of concomitance, while its changed and retained posture during motion can be seen as an instance of precursion. Therefore, it is possible that, when concomitance or precursion serves as a type of support relation that a co-event bears to the main motion event, languages are more likely to exhibit patterns deviating from their typological patterns in Talmyâs framework than when manner is involved.
This paper is organized as follows. Section 2 provides fundamental information on motion expressions in Kupsapiiny and Sidaama. Section 3 outlines the hypotheses addressed in this study. Section 4 describes the methodology. Section 5 discusses the syntax and the semantics of ideophones in Kupsapiiny and Sidaama. Section 6 presents video description data from speakers of these languages. Section 7 discusses the findings of this study. Section 8 concludes the paper.
In this paper, âmotionâ refers to âtranslational motionâ, which is the type of motion addressed in Talmyâs typology, unless otherwise specified as âself-contained motionâ. Dingemanseâs (2019: 16) cross-linguistic definition of an ideophone as âa member of an open lexical class of marked words that depict sensory imageryâ applies well to Kupsapiiny ideophones. However, it does not apply to Sidaama ideophones, which are âmembers of an open lexical class of marked wordsâ that do not necessarily depict sensory imagery but can, and which are conventionalized to the extent that they are not restricted to any specific semantic domain, such as one involving auditory or visual imagery.
For the transcription of Kupsapiiny data, this paper adopts a slightly modified version of the writing system developed for Bible translation by my main consultant, Chebet Francis, and his colleagues (Kupsapiiny Language Board & Kupsapiiny Language Development and Preservation Foundation, 2017), with added information on morpheme boundaries.2 For the transcription of Sidaama data, the International Phonetic Alphabet is used.
2 Background Information on Motion Expressions in Kupsapiiny and Sidaama
2.1 Background Information on Motion Expressions in Kupsapiiny
Kupsapiiny ([kÊ°ÊÌpsÉbɪÌ:ɲ]; also spelled Kupsapiny or Kupsabiiny) is a vso language, which has agglutinative morphology with a certain degree of fusion. It is predominantly an anti-verb-framed (non-head path-coding) language, which uses deictic directional affixes and path suffixes as satellites (Kawachi, 2025a). This language most frequently employs a single-verb satellite-framed construction (e.g., (1)) (Kawachi, 2021) as well as a non-verb-framed multi-verb construction where the main verb expresses a co-event such as a manner or cause of motion and the person-marked non-main verb expresses a path of motion (e.g., (2)).3 In (1), the manner verb raan âjumpâ takes the deictic-directionalââalongâ suffix complex -ÅÅn âcontinuously hitherâ and the path suffix -ÄÄ âalongâ (in this context). In (2), the main verb kÅny ârun (sg)â in its non-deictic-directional-affixed (ndda) form, which carries the tense-person prefix, is followed by the non-tensed, path verb leekitee âapproachâ, which carries only the subject person prefix.
Kupsapiiny also has verb-framed constructions (e.g., (3)), in which the main verb root expresses a path of motion, and the person-marked non-main verb or adverbial expresses a co-event, although these constructions are rarely used. In (3), the deictic-directional verb cÅÅ âcomeâ serves as the main verb, carrying the tense-person prefix, and precedes the non-tensed, manner verb (with the deictic-directionalââalongâ suffix complex -ÅÅn âcontinuously hitherâ and the imperfective aspect suffix -uu for âhitherâ).
Almost all motion verbs make a three-way directional-deixis distinction between their âhitherâ tonal-suprafix form, âthitherâ suffix form (with -t), and non-deictic-directional-affixed (ndda) form or a two-way directional-deixis distinction between their âhitherâ tonal-suprafix form and one of the other two forms. Table 1 illustrates these distinctions. (In this table, the forms are minimum citation forms, which are second-person singular present forms and are also used as imperatives. The final vowel suffix on the verb -u or -e marks the second-person subject for the verbâs class. The second-person singular prefix, which is a zero morpheme, is not represented.)
(4) provides the minimum citation forms of manner-of-motion verbs in this language. (For verbs with irregular plural forms, their plural forms, which include the second-person plural prefix a-, are additionally listed.)
There is also a verb, laany, that conflates a manner of motion and a path (âclimb upâ) when the figure entity is animate, or expresses only a path (âascendâ) when the figure entity is inanimate.
A manner of action (rather than a manner specifically of motion) can also be expressed with an adverbial, though there is no adverbial devoted to a manner of motion: e.g., lakwar âquicklyâ, mwayeek âfastâ, supiit âfastâ (< eng speed; sometimes preceded by the locative preposition Äm), mÅÅt âslowlyâ (often reduplicated), kapusa âhardâ (< swh). All of these can be reduplicated for emphasis. (None of them can directly follow the quotative marker unlike ideophones.)
This language uses the quotative marker kule (or its short form le, or less commonly a proximal demonstrative such as nyi âthisâ, cu âtheseâ, yuunyitÄnyi/yuunyitinyi âlike thisâ, or yuucitici/yuucitÄcu âlike theseâ) to quote speech (e.g., (5)) (Kawachi, Under review b). (As discussed in Section 5.1, one of the quotative markers accompanies ideophones.) There are many verbs that introduce reported speech with the quotative marker in any of the three speech report constructions, direct, indirect, and semi-direct speech. These include not only verbs of speech (e.g., mwoow-u âtell, sayâ, Åalaal âtalk, tellâ, tÄÄp âaskâ, kuur âcall (out)â, wÄÄc âreply, respondâ, caam âwhisperâ), but also verbs of actions that usually or often involve speech (e.g., kaast-e âthankâ, taÅaany âdisturbâ, neet âteachâ), verbs of cognition (e.g., nkÄt âknowâ, soot âthinkâ, kitee âwonderâ, kÄrÄsyi âlook carefullyâ), and verbs of emotion (e.g., faafan âbecome ashamedâ, muy âbecome afraid/shocked, nyokooriit-u âfearâ).
2.2 Background Information on Motion Expressions in Sidaama
Sidaama follows sov word order and exhibits aggultinative morphology with some fusion. It is a verb-framed language that often uses two types of mono-clausal converb constructions, the temporal overlap construction (toc) (e.g., (6)) and the temporal sequence construction (tsc) (e.g., (7)). These constructions use a toc converb and a tsc converb, respectively, to describe motion events that integrate their manner and path components into a single clause (Kawachi, 2025b). The converb and the main verb in each construction share the same subject. A toc converb takes the subject person-number(-gender) suffix, the infinitive suffix -a, and the ablative-instrumental-locative-manner suffix -nni (e.g., dod-ø-a-nni in (6)); -nni is often shortened to -i (e.g., dod-ø-a-i), and sometimes replaced by -a (e.g., dod-ø-a-a), which is a perseveration of the infinitive suffix -a.
Temporal overlap construction (toc)
In addition to the subject person-number(-gender) suffix, a tsc converb carries the connective converb suffix -e (e.g., dod-ø-e in (7)).
Temporal sequence construction (tsc)
The two constructions can be used interchangeably to describe the same motion event, though my consultants note that toc emphasizes the continuity of manner more strongly than tsc. Nevertheless, toc is more commonly used than tsc in descriptions of motion events that include the mention of manner (Kawachi, 2025b, see also Section 6.2).
When a manner verb and a path verb or one of the deictic-directional verbs (da- âcomeâ, haÉ- â(leave and) goâ, mar- âgo (and arrive)â) are used as V1 and V2 in the two constructions, both constructions function as core cosubordinations (Van Valin & LaPolla, 1997; Van Valin, 2005; Bohnemeyer & Van Valin, 2017). In this case, V1 and V2 obligatorily share operators at the level of the core juncture: both verbs must be under the scope of the deontic modal markers dand- âbe able to doâ and hasiis- âbecome necessaryâ.4 When other types of verbs are used, the constructions are less integrated and are interpreted as expressing separate eventsâtoc and tsc function as core (peripheral) subordination (âwhile doing V1, do V2â) and core coordination (âdo V1 and then do V2â), respectively. In tsc, the verbs can be modified by different temporal adverbials unlike toc, and the conjunction kaâentinni âthenâ or a tsc converb form of kaâ- âget up, finishâ (used to mean âthenâ) can intervene between them. Even when V1 and V2 are a manner verb and a path or deictic-directional verb, respectively, tsc can still allow an interpretation of separate events (e.g., âThe boy ran and entered the buildingâ in the case of (7)). This may explain why tsc is less preferred than toc for describing motion events when manner is mentioned as a co-event.
A manner verb, a path verb, and one of the deictic-directional verbs can co-occur in this order, with the deictic-directional verb as the main verb, in which case the path verb takes its tsc converb form. The tsc converb of the verb hiÉ¡- âreturn, turn, move toward, move via (lit., turn or direct oneâs body [toward])â often occurs anywhere in the middle of a sentence (sometimes in addition to another path verb) and functions like a directional postposition for the path âtowardâ (e.g., (16), (19)) or âviaâ.
(8) lists roots of manner-of-motion verbs in this language.
There are also verbs that express both a manner of motion and a path of motion: É¡idd- âclimb upâ, luutâ- âslide through a narrow space with difficultyâ, hulluukâ- âslide through a spaceâ. One of the deictic-directional self-agentive motion verbs haÉ- â(leave and) goâ, whose manner is usually interpreted as walking, can explicitly express this manner when used as a toc converb. Additionally, some verbs describe the motion of only a part of the figure entity rather than the whole figure entity: e.g., Å¡alakâ- âslipâ, kad- âtrample, tread onâ. When used as a toc converb, these verbs can express manners of continuous motion.
Manner-of-action verbs, rather than specifically manner-of-motion verbs, often occur as a tsc converb (rather than a toc converb) to express the manner of the motion of a motion verb functioning as the main verb: e.g., rak- âhurryâ, mudd-am- [rush-pass-] ârushâ, wiin-am- [rush-pass-] ârush (because of an urgent situation)â.
In this language, verb roots that are semantically dynamic rather than stative are classified into two types based on their aspectual properties, action verbs and state-change verbs (Kawachi, 2020: 546, 2023: 326). As Kawachi (2023: 326) summarizes, â[i]n the progressive aspect, an action verb expresses an ongoing action, but a change of state verb expresses a change of state in progress. An action verb does not combine with the continuous aspect, but a change of state verb does and denotes a continuous state up to the moment of speech. In the recent or distant perfect, an action verb expresses a completed action, whereas a change of state verb conveys a completed change of state or a current state as the result of the changeâ. Generally, manner-of-motion verbs fall into the category of action verbs, while posture-change verbs are classified as state-change verbs.
Sidaama also has an adverb for a manner of action in general (not restricted to a manner of motion), sununni/suutunni âslowlyâ, which can optionally be followed by a toc or tsc converb form of ikk- âbecomeâ with an intransitive main verb or ass- âdoâ with a transitive main verb. The language also has adverbs derived from adjectives through the addition of the clitic =É¡ede âlikeâ or the suffix complex -u-nni (-gen.m-ailm) (e.g., kaaÇ°Ç°a=É¡ede [strong/nice=like] âstrongly, nicelyâ, faayy-u-nni âin a pretty mannerâ [< faayya âprettyâ]). These adverbs can be used to describe a manner of action, though not specifically a manner of motion.
A verb denoting a change in the figure personâs physical sensation, psychological state, or posture can occur in its toc or tsc converb form with a motion verb as the main verb to express either a gradual ongoing change (e.g., bedeed-an-t-a-nni in (9); also, (20)) or the continued result of a change (e.g., Å¡okk-i-t-e in (10); also, (21), (22)).
The tsc converb of a transitive or causativized intransitive verb can be used with lekka âstep (lit., leg, foot)â or with this noun omitted to characterize the figure personâs steps as having a continued property resulting from a change in the steps while executing the motion of a main motion verb: e.g., bakkaaf- âstretch (oneâs legs) to take long stepsâ, halaÅ¡Å¡- âwidenâ (< halaâl- âbecome wideâ), harans- âshortenâ (< haraâm- âbecome shortâ), seed-i-s- [become.long-ep-caus-] âlengthenâ (< seed- âbecome longâ). This study treats this construction as expressing a posture change due to the verbsâ aspectual properties as state-change verbs.
Some verbs pertaining to speech require a tsc converb form of y- to precede them when quoting speech (e.g., (12)).
3 Hypotheses
Slobin (1997: 459, 2000: 111â115) argues that there are two types of manners of motion, basic and expressive manners of motion (or âtier oneâ and âtier twoâ manners). He proposes that, unlike satellite-framed languages, which have verbs for both types, verb-framed languages have verbs for basic manners (e.g., walk, fly), which are âneutral, everyday verbsâ, but do not have many verbs for expressive ones (e.g., dash, swoop, scramble), which are more âexceptionalâ (Slobin, 1997: 459).5 As a verb-framed language, Sidaama lacks expressive manner-of-motion verbs, as shown in (4) and (8). However, Kupsapiiny, which has satellites and can be classified as satellite-framed in single-verb constructions but never as verb-framed, also lacks expressive manner-of-motion verbs.
Schaefer (2001) shows that Emai (Edoid, Atlantic-Congo; Nigeria) (iso 639-3: ema), which he describes as a verb-framed language, uses ideophonic adverbs for expressive manners. In light of Slobinâs argument, he proposes the hypotheses that ideophones are largely restricted to verb-framed languages (also, Talmy in Ibarretxe-Antuñano 2005: 344), and that ideophones in these languages are used to depict expressive manners of motion, for which, according to Slobin (1997, 2000, 2004), there is a shortage of verbs in the lexical inventories.6 However, these hypotheses are based on Emai and a few other languages. Moreover, as Schaefer himself notes, ideophonic adverbs in Emai are not exclusively used to depict a manner of motion: they can modify not only verbs of motion but also verbs of state change, position change, and contact. Nevertheless, Schaefer frames these hypotheses within the context of Talmyâs motion typology. The present study considers these hypotheses worth testing and examines them using data from two typologically different languages.
A more general question is whether and how ideophones should be dealt with in Talmyâs (1985, 1991, 2000) motion typology. Talmy (in his comments in Ibarretxe-Antuñano, 2005: 344) states that the âsound-symbolic [m]anner-specifying constituent typeâ âseems not to be the sole (or perhaps even principal) locus for representing manner, but rather to be additional (and perhaps ancillary) to that of another constituent type: the more familiar type consisting of the gerundive or other non-finite form of a verb rootâ.
4 Methods
The present study analyzes the inventories of ideophones and manner-of-motion verbs in Kupsapiiny and Sidaama based on data collected from various sources. It also draws on the data that I collected using two sets of video stimulus elicitation tools: the video stimuli for the ninjal Project on Motion Event Descriptions across Languages (medal) (https://medalmotionevent.wordpress.com/), used in Matsumoto (2025), and the video stimuli for the Sound Symbolism for Locomotion (ssl, henceforth) project (https://osf.io/j9ab5/), used in Saji et al. (2019). In both elicitations, my main consultants instructed participants to watch video clips on our computer display and describe the event in each video clip in their language immediately after viewing it, as if answering the question âWhat happened?â (Ka-took-ø nee? (T.pst.3-happen-3 what(nom)) in Kupsapiiny and Ma-i ikk-ø-i? (what-nom.m happen-3sg.m-r.prf) in Sidaama). The video clips were played in randomized order for each participant. All the participants were accustomed to watching videos on tv and mobile devices and discussing the events depicted. After each session, I transcribed the recorded data with assistance from one of my main consultants. The present study analyzes the video elicitation data both quantitatively (by counting the frequency of ideophones used in descriptions of each stimulus set) and qualitatively (by examining the meanings of ideophones used to describe the video stimuli).
The medal video stimulus sets used in this study are stimulus sets A (data from 27 out of 52 clips) and C (44 clips), in which each clip depicts one of five manners of motionâwalking, running, skipping, jumping, or rollingâwithout sound, except for occasional audible footsteps.7 The author collected data from Kupsapiiny participants (stimulus set A: 12, stimulus set C: 17) and Sidaama participants (stimulus set A: 14, stimulus set C: 20) (see the introduction chapter of Matsumoto (2025) and the medal project website https://medalmotionevent.wordpress.com/ for details).
Saji et al.âs (2019) ssl motion video stimuli consist of 70 video clips, which each show a person moving from the left to the right of the display with an unusual manner of motion without producing any sound (see the ssl website https://osf.io/j9ab5/ for the clips).8 I collected data from five native speakers of each language using these clips.
5 Ideophones in Kupsapiiny and Sidaama
Childs (1994) provides an overview of ideophones in African languages but does not discuss Nilotic or Cushitic languages.
While no studies have examined ideophones in Nilotic or Cushitic languages within the context of motion typology, there are descriptions of ideophones in these language families. For example, Karani & Andrason (2022) describe ideophones in Arusa Maasai (Eastern Nilotic; Kenya; iso 639-3: mas), with a focus on the existence of both canonical and non-canonical (termed âprimaryâ and âgrammaticalizedâ in Dwyer & Moshi (2003)) ideophones in this language. There are also lists of ideophones in Cherangany (Kalenjin, Southern Nilotic; Kenya; closely related to Markweeta; [iso 639-3: enb]) by Mietzner (2016: 94â98) and in Luwo (Western Nilotic; South Sudan; iso 639-3: luo) by Storch (2014: 41â48). As for Cushitic, Treis (2008: 86â88, 2023, 2024) provides descriptions of ideophones in Kambaata (iso 639-3: ktb), another Highland Cushitic language of Ethiopia. Saeed (1999: 128â130) and Orkaydo (2013: 247â254) also list ideophones in Somali (Lowland East Cushitic; Somalia, Djibouti, and Ethiopia; iso 639-3: som) and Konso (Lowland East Cushitic; Ethiopia; iso 639-3: kxc), respectively.
5.1 Kupsapiiny Ideophones
Kupsapiiny ideophones must immediately follow one of the quotative markers mentioned in Section 2.1 (usually kule or its short form le) to form a quotative construction (Kawachi, Under review b). Unlike verbs, nouns, and adjectives, they never carry a tense prefix and cannot function as predicates. Together with the quotative marker, they occur only as an optional adjunct to the verb. These ideophones can express different types of concepts. Sound-mimetic ideophones are numerous and may be conventionalized or newly invented, whereas all other types of ideophones are relatively few and completely conventionalized.9 They are widely used in various situations and are not restricted to any specific register. The following descriptions first present ideophones used to describe motion events.
Some ideophones convey a manner of motion without expressing any sound, but only three such ideophones appeared in the data that the author reviewed with his consultants: lawus lawus (walking slowly in a very relaxed way with hands swinging without energy), nyankus nyankus (walking slowly in a relaxed way with the whole body swinging without energy), and wop wop (walking with hands swinging slowly).
When the main verb is a motion verb, ideophones mimic the sound that the figure physically produces (or the sound that the speaker perceives the figure as making or imagines them making). Most of these ideophones are repeated two, three, or four times (occasionally five times). For example, in (13), pata pata was used to describe the sound that the speaker thought the figure was making while running.
There are two major types of sound ideophones, sound-of-motion ideophones and sound-of-action ideophones. Sound-of-motion ideophones, which are devoted to motion events and are (not strictly but only loosely) associated with particular types of motion, express âproducing the sound while moving in a particular mannerâ (âas if the figure were producing the sound while moving in a particular mannerâ as used to describe the silent video clips). These ideophones generally represent sounds produced by the figureâs feet and are expressive in that their forms iconically reflect the depicted sounds. However, the manner-of-motion component encoded in sound-of-motion ideophones (e.g., in Table 2, âwalking slowly/tiptoeingâ in cakwas cakwas, âwalking in a hurryâ in cap cap) is not expressive. Some sound-of-motion ideophones also indicate a prototypical type of figure entity and/or a property of the ground entity. However, this information is cancellableâin fact, sound-of-motion ideophones can even be used for non-motion events as long as the form matches the sound that the speaker perceives. On the other hand, sound-of-action ideophones usually describe non-motion events by expressing âproducing the soundâ, but they appeared in participantsâ descriptions of silent motion video clips to convey âproducing the sound as if the figure were producing that soundâ. This suggests that the action component in their meaning is cancellable.
Tables 1 and 2 list all the sound-of-motion ideophones and sound-of-action ideophones that appeared in the participantsâ descriptions of the ssl video clips, along with their prototypical or representative uses. The forms shown are reduplicated once, though some were repeated more. The ideophones in Table 2, which occur with the quotative marker, accompany motion verbs, whereas those in Table 3 do not necessarily accompany motion verbs but may co-occur with verbs of any type of action or state change as long as the speaker perceives a corresponding the sound in the event (whether actually or fictively). While the ideophones in Tables 2 and 3 occurred with motion verbs in descriptions of the silent motion video clips, the participants who produced them perceived the figure as if they were producing those sounds. However, some video stimuli did not match the typical contexts in which the ideophones, especially those in Table 3, are usually used. This suggests that these ideophones primarily express sound rather than conforming strictly to their prototypical or representative situations.
Some sound-of-action ideophones were not used in the video descriptions at all (e.g., nyam nyam for eating, pu pu for fire burning, atisyoo for a sneeze, puu(u) for a solid object falling into a liquid). Some of these, especially those that express the sound of a punctual event, are not reduplicated, though they may be repeated if the sound is perceived as recurring.
Additionally, three other types of ideophones occur with the quotative marker, though none appeared in the video descriptions. First, some ideophones describe animal sounds (e.g., mpoo(o) for a cow bellowing, mee(e) for a goat bleating).
Second, there are manner-of-action or state-change ideophones that mostly express completeness. These are not sound-mimetic and have no meaning outside specific contexts. (14) lists such ideophones, each occurring with the quotative marker kule. Most of these collocate with only a limited set of preceding verbs. These ideophones are not reduplicated.
Third, there are color ideophones, as shown in (15). Each of these must be used with the word for a specific color preceding the quotative marker to form an idiomatic expression. They cannot occur in any other context and are not reduplicated.
5.2 Sidaama Ideophones
Sidaama ideophones are more grammaticalized (Dwyer & Moshi, 2003; Andrason & Heine, 2023), more deideophonized, and less expressive (Dingemanse & Akita, 2016; Dingemanse, 2017) than Kupsapiiny ideophones. Like ideophones in other Ethiopian languages (Ferguson, 1970; Cohen et al., 2002), Sidaama ideophones are idiomatized and are accompanied by the verb of saying y- or the verb of doing ass- (e.g., (16)), whose tsc converb form functions as a quotative marker in direct speech. Most ideophones with y- and with ass- show contrast in transitivity: Those with y- are intransitive and those with ass- are transitive or causative. Nearly all ideophones must be followed by one of these verbs and cannot occur outside of these idioms, except for iÅ¡Å¡i y- âagreeâ (ass-: transitive/causative), which contains the word for âyesâ in Amharic. Many of the ideophones end in i, and a small number of them end in e, a, o, or u. The ideophones have high pitch on their penultimate vowels.
There are no morphosyntactic restrictions on the idioms. Both y- and ass- can take any suffix, and the idioms can function as main verbs, converbs, or verbs in subordinate clauses.
Sidaama ideophones are not restricted to any specific semantic domain: they can express manners (in general or of motion), paths of motion, state changes, posture changes, perception, emotion, or sensation. Most of them are not onomatopoeic. Ideophones are not productive, and novel ideophones are difficult to create. There are approximately 100 ideophones, and those used to describe motion events are highly restricted in number. Therefore, they do not fill the gap in the small inventory of manner-of-motion verbs (especially expressive ones) in (8) in this v-language.
(17) lists motion ideophone idioms: (17a) and (17b) express a path and a manner of motion, respectively. None of the manner-of-motion ideophone idioms in (17b) conveys any expressive manners of motion. There is also one idiom that expresses both: Å¡urri y- âwalk fast along a narrow pathâ (ass-: transitive/causative). (The ideophones that must be repeated are shown as repeated, and those that are optionally repeated are in parentheses.)
These path and manner ideophone idioms syntactically behave as path-of-motion and manner-of-motion verbs, as described in Section 2.2. For example, Å¡ikkâi y- âapproachâ, which follows the manner verb dod- ârunâ in (18) and (19), is used in its tsc converb form in (18) and as a main verb in (19).
There is also an idiom that expresses only the fact of motion, milli y- âmoveâ, whose reduplicated form milli milli y- can mean âmove repeatedlyâ or the self-contained motion of âwiggleâ. In describing the ssl video clips, two out of the five participants used the transitive/causative counterpart of this idiom milli ass- and its reduplicated form milli milli ass- to mean âmove (oneâs body part) (repeatedly)â.
Some ideophone idioms convey a manner of action in general, and their tsc converb communicates a manner of motion when paired with a motion verb as the main verb, like manner-of-action verbs (Section 2.2): e.g., kâuuttâi y- âhurryâ (ass-: transitive/causative).
Some ideophone idioms describe a posture change, as in Table 4.
When the main verb is a motion verb, posture-change ideophones with a toc converb form of y-/ass- express a gradual ongoing posture change (âmove, changing oneâs posture graduallyâ) (e.g., (20)) and those with a tsc converb form of y-/ass- express a changed and retained posture (âchange oneâs posture and move in the postureâ) (e.g., (21), (22)).
There are also ideophones for other semantic domains, as in Table 5. Some of these, which convey the meaning of a state change with a toc and tsc converb form of y-/ass-, also express a gradual, ongoing state change (e.g., leÄÄâi y-i-t-a-nni used instead in (23)) and a changed and retained state (e.g., (23)), respectively, when combined with a motion verb as the main verb.
There are onomatopoeic ideophones, as shown in Table 6.
About half of the idioms with ass- have transitive verbs ending in -eess or -is, into which they seem to have developed. (Also, about half of the ideophones have nouns nominalized from them.) Most verbs ending in -eess and those ending in -is have the same meaning, but a small number of verbs ending in -is (e.g., meddis-) are intransitive and have the same meaning as their y- idiom counterparts (e.g., meddi y- âbe about to arriveâ). Such idioms are not restricted to any semantic domain. Table 7 lists verbs that seem to have derived from some of the idioms in (17).
5.3 Findings on Manner-of-Motion Verbs and Ideophones in Kupsapiiny and Sidaama
As shown so far, Kupsapiiny and Sidaama have 20 (18 if two pairs of exact synonyms are lumped together) and 14 manner-of-motion verbs (see (4) and (8)), respectively. Neither language has verbs for particularly expressive manners of motion. The number of manner-of-motion verbs is small. See Slobin, 1997: 458â460, 2000: 113â115, 118â123, 134, 2006: 71, and Matsumoto (Forthcoming) for a comparison of inventories of manner-of-motion verbs across verb-framed and satellite-framed languages. Slobin (2006: 71) reports that â[t]he satellite-framed languages that [he has] examinedâEnglish, German, Dutch, Russian, and Hungarianâeach have several hundred manner verbsâ, whereas verb-framed languages such as âSpanish, French, Turkish, and Hebrew have less than 100, and probably less than 60 in everyday useâ.
Contrary to Schaeferâs (2001) hypothesis that ideophones are mostly limited to verb-framed languages, Kupsapiiny has many expressive ideophones that can be used in motion event descriptions, unlike Sidaama, though the high degree of expressiveness is found in the sound, rather than the manner of motion, conveyed by Kupsapiiny sound-of-motion ideophones.10 Therefore, for these languages, it is questionable whether ideophones depict expressive manners of motion or contribute to Slobinâs second-tier manner lexicon.
6 Results of Video Description Data
This section presents data on the frequency of the use of ideophones in the two languages to describe the video clips in the three sets of video stimuli. In descriptions of the two sets of medal stimuli, ideophones did not occur in Kupsapiiny and were rarely used in Sidaama. In contrast, in descriptions of the ssl stimuli, ideophones were produced, but their frequency differed between the two languages: it was much higher in Kupsapiiny (88.3%) than in Sidaama (24.9%). In both languages, ideophones occurred as optional elements due to their grammatical properties in Kupsapiiny and despite their potential to serve as the main verb in Sidaama. In the Kupsapiiny descriptions with ideophones, the following verbs were used as the main verb: wÅ âgoâ, kÅny ârun (sg)â, lapat ârunâ, wÅsyÄÄt/wÄsyÄÄt âwalk (sg)â, wÅÅstÅÅte/wastaate âwalk (sg)â. In the Sidaama descriptions with ideophones, the following verbs occupied the main verb position: haâr- â(leave and) goâ (usually interpreted as âwith the manner of walkingâ), saâ- âpassâ, dod- ârunâ, kâaaf- âstride, take (long) steps, step overâ, yaas- âproceed slowly and quietlyâ, kad- âtrampleâ, kubb- âjumpâ.
6.1 Occurrence of Ideophones in the Kupsapiiny Data
No ideophones appeared in the medal video elicitations from Kupsapiiny speakers. In contrast, in describing the ssl video clips, the Kupsapiiny participants used ideophones with the quote marker (kule or its short form le) in many cases: 309 out of 350 responses (88.3%), as shown in Table 8. Therefore, this language generally uses ideophones in descriptions of motion events involving unusual manners. However, the ideophones used by the participants were not necessarily the three manner-of-motion ideophones (65 responses: 21.0%) mentioned in Section 5.1. The remaining ideophones were sound-of-motion ideophones (198) and sound-of-action ideophones (46). Thus, even though no sound was heard in the events depicted in the video clips, 244 (79.0%) of the ideophones used were those expressing sound, with 198 of them representing the sound of motion, which were used to describe the sound that the speaker perceived the moving figure was likely making. While sound ideophones can convey the occurrence of a sound in an expressive way, none of them communicates the manners of motion that they are associated with in an expressive way (Section 5.1). Table 8 shows how many times each participant used the three types of constituents as a non-main verb.
6.2 Occurrence of Ideophones in the Sidaama Data
In the medal video elicitations from Sidaama speakers, an ideophone appeared as part of a motion description twice in the speakersâ descriptions of video stimulus set A (out of 378 descriptions) and once in the descriptions of video stimulus set C (out of 880 descriptions). A tsc converb of kâassi y- âbecome quietâ (< amh) occurred, as in (16), and Å¡ikkâi y- âapproachâ appeared both in its tsc converb form, as in (18), and as a main verb, as in (19). None of the manner-of-motion ideophones in (17b) occurred, possibly because none was suitable for any of the events in the videos. One ideophone was used to descibe the figure personâs posture change of turning back, whose resultative state can be interpreted as being maintained by the figure, as in (21).
Sidaama ideophones were used for some of the ssl video stimuli. However, their frequency was much lower than that of Kupsapiiny ideophones: only 87 times out of the total 350 responses (24.9%)âor 82 responses out of 350 responses (23.4%) if counting only the number of responses, as five responses contained two ideophones. The verb y-/ass- following the ideophones never occurred as a main verb, though it could have, as in (19), but always appeared as a toc or tsc converb. Table 9 shows the frequency of the use of the ideophones as converbs in these two constructions. Among the manner-of-motion ideophones in (17b), only the transitive/causative form of Å¡irri y- âslideâ occurred just twice to describe the figure person sliding her legs, though a few others could have been used to describe events in the videos. The idiom milli (milli) ass- âmove (repeatedly or continuously) (transitive)â was used 41 times (as in Table 10; 36 times by Participant #1 and 3 times by Participant #3) to describe the figure personâs hand or leg movement, without specifying the manner. This was analyzed as an expression of the fact of motion, though it could also be interpreted as self-contained motion or possibly as a manner of motion. The item immediately below milli milli ass- in Table 9, leÄÄâi y-, denotes a change in physical sensation âbecome tiredâ and was used as in (23). All the other idioms were those expressing a type of posture change.
Posture changes are not necessarily expressed with ideophones but can also be encoded in posture-change verbs (or through a combination of transitive verbs, especially those expressing posture changes, and body-part object noun phrases). As shown in Table 10, posture-change verbs (and their objects) occurred 170 times. In 205 out of 350 responses (58.6%) for the ssl video clips, a posture change was mentioned at least once (Participant #1: 44, #2: 46, #3: 40, #4: 42, #5: 33), whereas in 145 out of 350 responses (41.4%), no posture change was mentioned (Participant #1: 26, #2: 24, #3: 30, #4: 28, #5: 37). Additionally, the stimuli were described with a manner-of-action adverbial 51 times and with a manner-of-motion or manner-of-action verb as a converb 147 times. This suggests that Sidaama speakers often conceptualize motion events with different unusual manners not only as those involving manners but also as those involving posture changes.
Unlike milli (milli) ass-, which functions as an action verb rather than a state-change verb, the other idioms used by the participants act as state-change verbs, whose progressive form expresses a gradual state change (Section 2.2). Posture-change idioms occurred as tsc converbs 33 times and as toc converbs 10 times. Similarly, posture-change verbs also appeared more frequently as tsc converbs (105 times) than as toc converbs (65 times).
tsc is preferred over toc because the posture-change idioms and the posture-change verbs as tsc converbs express the retention of a changed posture (âchange oneâs posture and move in the postureâ), unlike those as toc converbs, which convey a gradual posture change in the process of motion (âmove, changing oneâs posture graduallyâ). This differs from manner-of-motion verbs, which appeared more frequently as toc converbs (156 instances) than as tsc converbs (12 instances) (both toc and tsc: 2 instances) in the participantsâ descriptions of medal stimulus set A (Kawachi 2025b).
7 Discussion
The Kupsapiiny data and the Sidaama data each reject Schaeferâs (2001) hypothesis that ideophones are found mostly in verb-framed languages, as well as his hypothesis that ideophones are used to convey expressive manners of motion, for which Slobin (1997, 2000, 2004) postulates verb-framed languages lack verbs in their lexicons. Neither Kupsapiiny, which uses satellites, nor Sidaama, which is verb-framed, possesses expressive manner-of-motion verbs. Kupsapiiny has a large number of ideophones, most of which depict sounds expressively, while some describe manners of motion only in an abstract way. Sidaama has only five manner-of-motion ideophones, few of which are particularly expressive.
In Kupsapiiny, the quotative marker and an ideophone together always function as optional elements due to their grammatical properties. Also in all Sidaama video descriptions, the idiom consisting of an ideophone and y-/ass- occurred but only as a converb, never as a main verb, although it could occur as one. Therefore, unlike ideophones with the quotative marker in Kupsapiiny, manner-of-motion ideophone idioms in Sidaama behave syntactically similarly to manner-of-motion verbs. If Sidaama had a large inventory of ideophones for expressive manners of motion, they might be analyzed as substitutes for expressive manner-of-motion verbs in its lexical inventories, but it does not.
At least in Kupsapiiny and Sidaama, the relative inventory size of ideophones versus expressive manner-of-motion verbs seems to influence the use of ideophones. This raises the question of how ideophones should be situated within Talmyâs motion typology. To explore this, the present study first examines the types of support relations that the co-events bear to the main motion event in the semantics of Kupsapiiny and Sidaama ideophones, following the approach of Toratani (2012) for Japanese ideophones.
the [c]o-event co-occurs with the [m]otion event and is conceptualized as an additional activity that the [f]igure of the [m]otion event exhibitsâan activity that directly pertains to the [m]otion event but that is distinct from it. In this conceptualization, the [c]o-event can âpertainâ to the [m]otion event in several ways, such as by interacting with it, affecting it, or being able to manifest itself only in the course of it. Thus, the [c]o-event can consist of a pattern of motion by the [f]igureâspecifically, a so-conceivedly abstractable type of self-contained motionâthat coalesces with the [f]igureâs translational motion to form a more complex envelope of movement, ⦠[o]r the [c]o-event can be a conceptually abstractable activity by the [f]igure that could exist only in association with translational motion by the [f]igure.
Neither the sounds that many Kupsapiiny ideophones describe as being produced (e.g., (13)) nor the posture changes and their resultative postures represented by some ideophone idioms and verbs in Sidaama (e.g., (20), (22)) satisfy this definition of the co-event of manner or qualify as manners of motion. The events that they express do not directly âpertainâ to the main motion event; rather, they are perceived as more separable from and more independent of it than manners of motion, as they could occur even if the main motion event did not take place.
the [c]o-event co-occurs with the main [m]otion event and is an activity that the [f]igure of the [m]otion event additionally exhibits. But here, this activity does not in itself pertain to the concurrent [m]otion, in the sense of âpertainâ just described, and could just as readily take place by itself (although the presumed difference between [m]anner and [c]oncomitance may have the character more of a gradient than of a sharp division).
Resultative postures expressed by the tsc converbs of Sidaama ideophones and verbs (e.g., (22)) exhibit a support relation of precursion with the main motion event component, in which âthe [c]o-event precedes the main [m]otion event but does not cause or assist its occurrence. The [m]otion event would proceed much the same if the [c]o-event had not occurredâ (Talmy, 2000: 42).
The co-events of concomitance and precursion (and the co-event of subsequence) are not as conceptually and semantically linked to the framing event of motion as the co-event of manner of motion. Consequently, ideophonic elements used for the co-events of concomitance and precursion are likely to be less syntactically integrated with the main motion verb than manner-of-motion verbs. This allows Kupsapiiny sound-of-motion or sound-of-action ideophones with the quotative marker as well as Sidaama posture-change ideophones with y-/ass- (and also posture-change verbs) to modify manner-of-motion verbs, as in (13) and (20)/(22), respectively.
Talmyâs motion typology as part of his typology of event integration has long been examined primarily for motion events whose co-event is a manner of motion (e.g., Slobin, 2006), but rarely for those involving other types of co-events (with the exceptions of, for example, Kawachi, 2012, 2016, 2018). Therefore, it is worth exploring how Talmyâs typology applies to events with other types of co-events.
Talmy (2000, 2017) provides the English sentences in (24) and (25) as examples of the support relations of subsequence and precursion, respectively. (Unlike (25), which contains the satellite on, (24a) and (24b) do not include any satellite and therefore are not instances of a satellite-framed construction, but rather belong to the type of construction typically found in satellited-framed languages. Nevertheless, the prepositional phrases across the sky and to the party could be replaced by satellites.)
Some verb-framed languages seem to express motion events with these support relations in a less integrated way than those with the support relation of manner (Kawachi, 2018). For example, Sidaama uses the core (peripheral) subordination toc subconstruction (âwhile doing V1, do V2â) for (24a), and the core coordination tsc subconstruction (âdo V1 and then do V2â) for (24b) and (25) (Section 2.2). Sidaama posture-change ideophones with y-/ass- and posture-change verbs often occur as the converb in one of these constructions, as they always did in the participantsâ descriptions of the ssl motion video stimuli.
Even in languages with path satellites, (some of) the satellites can or cannot be combined with a non-motion verb to form a single clause. (26a2), (26b), and (27) show Kupsapiiny approximate equivalents of (24a), (24b), and (25), respectively.
(26a2), (26b), and (27) use associated motion (am) constructions (Koch, 1984; Wilkins; 1991; Guillaume, 2013, 2016; Guillaume & Koch, 2021), which are grammatical devices that motionalize non-motion verbs by specifying the path (particularly direction) of the motion component. Specifically, (26a2) and (26b) are instances of the deictic-directionalââalongâ suffix complex construction (Kawachi, 2021) and (27) is an instance of the applicative suffix construction. Both constructions can be used for concurrent am (e.g., âeat while goingâ) or subsequent am (e.g., âeat then goâ) (Kawachi, Under review a), though the applicative suffix construction is restricted to state-change verbs. Therefore, languages with grammatical am markers can relatively easily express motion events involving these support relations using a satellite-framed construction, in which the non-motion event component appears in the verb root, while the path-of-motion component is realized in the morphological am marker, which functions as a path satelliteâprovided that a verb root can be used for the non-motion event component.
On the other hand, (26a1) is more complex than the English sentence (24a). Kuprapinny can express the specific path for âacrossâ only with the verb mur âcrossâ (kurÅÅsiinkÄn âcrossâ (< eng cross), muut âcut, crossâ, or yaay âcrossâ). Therefore, whether languages with grammatical am markers can use a compact construction depends on the type or complexity of the path in the described motion event, compared to the expressible paths of am markers and other path satellites. Such languages can follow the satellite-framed pattern as long as their am markers can be used for the paths they describe, and the non-motion event component, which has the support relation of concomitance or precursion (or that of subsequence) with the motion component, is expressed in a verb root.
the verb expresses at once both the fact of [m]otion and a [c]o-event, usually either the manner or the cause of the [m]otion. A language of this type has a whole series of verbs in common use that express motion occurring in various manners or by various causes (2000: 27).
In fact, manner seems more likely to appear in the verb root than concomitance or precursion (or subsequence). For these types of co-events to occur in the verb root, the language must have morphological am markers (or syntactically compact constructions, as in the English constructions in (24a), (24b), and (25)). However, even languages with such markers cannot use them when they cannot express the path to be described with them or when they do not have a verb that can convey the co-event, as in Kupsapiiny, which lacks a verb for any of the sounds expressed with sound ideophones.
8 Conclusion
Using data from two typologically different languages, Kupsapiiny and Sidaama, this paper has tested Schaeferâs (2001) hypothesis that ideophones are primarily found in verb-framed languages, as well as his hypothesis that ideophones are used to convey expressive manners of motion, for which Slobin (1997, 2000, 2004) argues verb-framed languages lack sufficient verbs. The data from both Kupsapiiny and Sidaama contradict these hypotheses: ideophones for manners of motion are not predominantly found in verb-framed languages, nor can the inventory of ideophones be linked to a lack of manner-of-motion verbs in verb-framed languages. Kupsapiiny, an anti-verb-framed language, prefers to use satellites or non-main verbs to express a path of motion but has a large number of ideophones, many of which depict sounds expressively. Although Sidaama is verb-framed, it has only a limited number of manner-of-motion ideophone idioms, as seen in (17b). This language tends to describe events with unusual manners of motion using ideophone idioms, particularly those involving posture changes, for which verbs are also available. The data from Kupsapiiny and Sidaama suggest that speakers of different languages can conceptualize motion events with manner as those involving another type of co-event (precursion or concomitance) and describe them in a more foregrounded way than manners of motion. Moreover, the elements expressing them are syntactically more loosely linked with a main verb for the main motion event than those expressing manner of motion. In these respects, the ideophones in these languages function as an âadditional (and perhaps ancillary)â (Talmyâs comments in Ibarretxe-Antuñano, 2005: 344) constituent type.
Research on ideophones in the context of Talmyâs motion typology has long assumed that ideophones used in motion event descriptions express manners of motion (e.g., Slobin, 2004; Ibarretxe-Antuñano, 2019). However, the findings of the present study suggest that the motion component that speakers of (at least some) languages conceptualize with certain types of ideophones is not a manner of motion as defined by Talmy (2000: 45). Ideophones may appear to express manners of motion when some languages use them to describe motion events that other languages encode with manner-of-motion verbs, but this interpretation arises only when viewed from the perspective of the latter type of language. Ideophones and manner-of-motion verbs differ morpho-syntactically and the types of motion components that they express can also differ.
Abbreviations
| accobl | accusative-oblique |
| ailm | ablative-instrumental-locative-manner |
| al | âalong, continuouslyâ |
| all | allative |
| amh | Amharic |
| caus | causative |
| cmpl | complementizer |
| cnn | connective converb |
| d.prf | distant perfect |
| datloc | dative-locative |
| dm | discourse marker |
| eng | English |
| ep | epenthesis |
| F | feminine |
| gen | genitive |
| H | âhitherâ |
| impers | impersonal |
| inf | infinitive |
| ipfv | imperfective |
| lv | lengthened vowel |
| M | masculine |
| ndda | non-deictic-directional-affixed |
| nom | nominative |
| poss | possessive |
| prog | progressive |
| prs | present |
| quot | quotative marker |
| r.prf | recent perfect |
| ref | reflexive |
| rel | relative |
| sg | singular |
| swh | Swahili |
| t.pst | today past |
| th | âthitherâ |
| vafw | âvia (along, through, over, past, etc.), at, from, with/usingâ |
Acknowledgments
I would like to express my deepest thanks to my primary Kupsapiiny consultants, Chebet Francis and Chebet Mercy, and my main Sidaama consultants, Genene Gudura and Iyasu Gudura, for their invaluable assistance over the years. My heartfelt thanks also go to the participants in the video description tasks for providing the data. I am profoundly grateful to Kiyoko Toratani and Kimi Akita, as well as to the anonymous reviewers, for their detailed and insightful comments on the manuscripts. This study was supported by jsps kakenhi (15H05157, 19K00565, 22K18479, 22H00659, 23K21931; pi: Kazuhiro Kawachi), jsps kakenhi (15H03206, 19H01264; pi: Yo Matsumoto), jsps kakenhi (16H01928; pi: Mutsumi Imai), the Keio Gijuku Fukuzawa Memorial Fund for the Advancement of Education and Research, the Keio University Academic Development Funds, the Ushioda Memorial Fund, and ninjal (âEvidence-based Theoretical and Typological Linguisticsâ).
References
Ameka, Felix K., and James Essegbey. 2006. Elements of the grammar of space in Ewe. In Stephen C. Levinson and David P. Wilkins (eds.), Grammars of Space: Explorations in Cognitive Diversity, 359â399. Cambridge: Cambridge University Press.
Andrason, Alexander, and Bernd Heine. 2023. On the grammaticalization of ideophones. In Sylvie Hancil and Vittorio Tantucci (eds.), Different Slants on Grammaticalization, 237â262. Amsterdam/Philadelphia: John Benjamins.
Bohnemeyer, Jürgen & Robert D. Van Valin, Jr. 2017. The macro-event property and the layered structure of the clause. Studies in Language 41: 142â197. https://doi.org/10.1075/sl.41.1.05van.
Childs, G. Tucker. 1994. African ideophones. In Leanne Hinton, Johanna Nichols, and John J. Ohala (eds.), Sound Symbolism, 178â204. Cambridge, MA: Cambridge University Press.
Childs, G. Tucker. 2003. An Introduction to African Languages. Amsterdam/Philadelphia: John Benjamins.
Cohen, David, Marie-Claude Simeone-Senelle, and Martine Vanhove. 2002. The grammaticalization of âsayâ and âdoâ: An areal phenomenon in East Africa. In Tom Güldemann and Manfred von Rocador (eds.), Reported Discourse: A Meeting Ground for Different Linguistic Domains, 227â251. Amsterdam/Philadelphia: John Benjamins.
Dingemanse, Mark. 2017. Expressiveness and system integration: On the typology of ideophones, with special reference to Siwu. stufâ Language Typology and Universals 70: 363â384. https://doi.org/10.1515/stuf-2017-0018.
Dingemanse, Mark. 2019. âIdeophoneâ as a comparative concept. In Kimi Akita and Prashant Pardeshi (eds.), Ideophones, Mimetics, and Expressives, 13â33. Amsterdam: John Benjamins Publishing Company.
Dingemanse, Mark, and Kimi Akita. 2016. An inverse relation between expressiveness and grammatical integration: On the morphosyntactic typology of ideophones, with special reference to Japanese. Journal of Linguistics 53: 501â532. https://doi.org/10.1017/S002222671600030X.
Dwyer, David, and Lioba Moshi. 2003. Primary and grammaticalized ideophones. In John M. Mugane (ed.), Linguistic Typology and Representation of African Languages, 173â185. Trenton: Africa World Press.
Ferguson, Charles. 1970. The Ethiopian language area. Journal of Ethiopian Studies 8: 67â80. (also published in 1976 in M. Lionel Bender, J. Donald Bowen, Robert Cooper, and Charles Ferguson (eds.), Language in Ethiopia, 63â76. Oxford: Oxford University Press.).
Guillaume, Antoine. 2013. Reconstructing the category of âassociated motionâ in Tacanan languages (Amazonian Bolivia and Peru). In Ritsuko Kikusawa and Lawrence A. Reid (eds.), 2013. Historical Linguistics 2011: Selected Papers from the 20th International Conference on Historical Linguistics, Osaka, 25â30 July 2011, 129â151. Amsterdam: Benjamins.
Guillaume, Antoine. 2016. Associated motion in South America: Typological and areal perspectives. Linguistic Typology 20: 81â177. https://doi.org/10.1515/lingty-2016-0003.
Guillaume, Antoine, and Harold Koch. 2021 (eds.), Associated Motion. Berlin & Boston: De Gruyter Mouton. https://doi.org/10.1515/9783110692099.
Ibarretxe-Antuñano, Iraide. 2005. Leonard Talmy. A windowing to conceptual structure and language. Part 1: Lexicalization and typology. Annual Review of Cognitive Linguistics 3: 325â347. https://doi.org/10.1075/arcl.3.17iba?locatt=mode:legacy.
Ibarretxe-Antuñano, Iraide. 2019. Towards a semantic typological classification of motion ideophones: The motion semantic grid. In Kimi Akita and Prashant Pardeshi (eds.), Ideophones, Mimetics and Expressives, 137â166. Amsterdam/Philadelphia: John Benjamins.
Karani, Michael, and Alexander Andrason. 2022. Ideophones in Arusa Maasai: Syntax, morphology, and phonetics. Open Linguistics 8: 440â458. https://doi.org/10.1515/opli-2022-0220.
Kawachi, Kazuhiro. 2012. Event integration patterns in Sidaama (Sidamo). Proceedings of the 34th Annual Meeting of the Berkeley Linguistics Society, General Session, 175â186. https://doi.org/10.3765/bls.v34i1.3566.
Kawachi, Kazuhiro. 2016. Introduction: An overview of event integration patterns in African languages. Asian and African Languages and Linguistics 10: 1â36. https://doi.org/10.15026/85062.
Kawachi, Kazuhiro. 2018. Event integration patterns in Sidaama and Japanese. In Prashant Pardeshi and Taro Kageyama (eds.), Handbook of Japanese Contrastive Linguistics, 313â342. Berlin/Boston: De Gruyter Mouton. https://doi.org/10.1515/9781614514077-012.
Kawachi, Kazuhiro. 2020. Chapter 42: Sidaama. In Rainer Vossen and Gerrit J. Dimmendaal (eds.), The Oxford Handbook of African Languages, 543â552. Oxford: Oxford University Press. https://doi.org/10.1093/oxfordhb/9780199609895.013.30.
Kawachi, Kazuhiro. 2021. The âalongââdeictic-directional verb suffix complex in Kupsapiny. In Antoine Guillaume and Harold Koch (eds.), Associated Motion, 747â777. Berlin & Boston: De Gruyter Mouton. https://doi.org/10.1515/9783110692099-019.
Kawachi, Kazuhiro. 2023. Chapter 17: Sidaama. In Ronny Meyer, Bedilu Wakjira, and Zelealem Leyew (eds.), The Oxford Handbook of Ethiopian Languages, 320â343. Oxford: Oxford University Press. https://doi.org/10.1093/oxfordhb/9780198728542.013.17.
Kawachi, Kazuhiro. 2025a. Kupsapiiny as a predominantly anti-verb-framed, nonhead path-coding language. In Yo Matsumoto (ed.), Motion Event Descriptions across Languages, Volume 1: Case Studies of Linguistic Representations of Motion, 91â128. Berlin & Boston: De Gruyter Mouton. https://doi.org/10.1515/9783110690866-003.
Kawachi, Kazuhiro. 2025b. Does the degree of integration of constructions iconically reflect motion event types? The case of Sidaama. In Yo Matsumoto (ed.), Motion Event Descriptions across Languages, Volume 1: Case Studies of Linguistic Representations of Motion, 129â166. Berlin & Boston: De Gruyter Mouton. https://doi.org/10.1515/9783110690866-004.
Kawachi, Kazuhiro. Under review a. Degrees of grammaticalization and crosslinguistic implicational hierarchies: An investigation of the associated motion constructions in Kupsapiiny.
Kawachi, Kazuhiro. Under review b. Speech reports in Kupsapiiny.
Kawachi, Kazuhiro & Ikuko Matsuse. Under review. High frequency of mention of deictic direction in Kupsapiiny and Kathmandu Newar.
Killian-Hatz, Christa. 2001. Universality and diversity: Ideophones from Baka and Kxoe. In F. K. Erhard Voeltz and Christa Kilian-Hatz (eds.), Ideophones, 155â163. Amsterdam/Philadelphia: John Benjamins.
Koch, Harold. 1984. The category of âassociated motionâ in Kaytej. Language in Central Australia 1: 23â34.
Kupsapiiny Language Board and Kupsapiiny Language Development and Preservation Foundation (2017 [2011]) An Orthography and Spelling Guide for Kupsapiiny, Revised version. Entebbe, Uganda: sil International.
Matsumoto, Yo. (ed.) 2025. Motion Event Descriptions across Languages, Volume 1: Case Studies of Linguistic Representations of Motion. Berlin & Boston: De Gruyter Mouton. https://doi.org/10.1515/9783110690866 (ninjal Motion Event Descriptions across Languages project website: https://medalmotionevent.wordpress.com/ for details).
Matsumoto, Yo. (ed.) Forthcoming. Motion Event Descriptions across Languages, Volume 2: Theory and Typology. Berlin & Boston: De Gruyter Mouton.
Matsuse, Ikuko. 2025. Motion event descriptions and deictic verbs in Kathmandu Newar: An experimental study. In Yo Matsumoto (ed.), Motion Event Descriptions across Languages, Volume 1: Case Studies of Linguistic Representations of Motion. Berlin & Boston: De Gruyter Mouton. https://doi.org/10.1515/9783110690866-016.
Mietzner, Angelika. 2016. Cherangany: A Kalenjin Language of Kenya. Cologne: Rüdiger Köppe.
Nakagawa, Hirosi. 2011. A first report on G|ui ideophones. In Osamu Hieda, Christa König, and Hirosi Nakagawa (eds.), Geographical Typology and Linguistic Areas: with Special Reference to Africa, 279â286. Amsterdam/Philadelphia: John Benjamins.
Nakagawa, Hirosi. 2024. Onomatopoeia in GÇui (Kalahari Khoe). In LÃvia Körtvélyessy and Pavol Å tekauer (eds.), Onomatopoeia in the Worldâs Languages: A Comparative Handbook, 197â207. Berlin & Boston: De Gruyter Mouton.
Ohara, Kyoko Hirose. 2002. Linguistic encodings of motion events in Japanese and English: A preliminary look. The Hiyoshi Review of English Studies 41: 122â153. Keio University. https://koara.lib.keio.ac.jp/xoonips/modules/xoonips/detail.php?koara_id=AN10030060-20020930-0122.
Orkaydo, Ongaye Oda. 2013. A grammar of Konso. PhD thesis, Leiden University. https://www.lotpublications.nl/Documents/326_fulltext.pdf.
Saeed, John. 1999. Somali. Amsterdam/Philadelphia: John Benjamins.
Saji, Noburo, Kimi Akita, Katerina Kantartzis, Sotaro Kita, and Mutsumi Imai. 2019. Cross-linguistically shared and language-specific sound symbolism in novel words elicited by locomotion videos in Japanese and English. PLoS ONE 14.7, e0218707. https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0218707 (Sound Symbolism for Locomotion project website: https://osf.io/j9ab5/).
Samarin, William. 1970. Inventory and choice in expressive language. Word 26: 153â169.
Schaefer, Ronald P. 2001. Ideophonic adverbs and manner gaps in Emai. In F. K. Erhard Voeltz and Christa Kilian-Hatz (eds.), Ideophones, 339â354. Amsterdam/Philadelphia: John Benjamins.
Slobin, Dan I. 1997. Mind, code and text. In Joan L. Bybee, John Haiman, and Sandra A. Thompson (eds.), Essays on Language Function and Language Type: Dedicated to T. Givón, 437â467. Amsterdam/Philadelphia: John Benjamins.
Slobin, Dan I. 2000. Verbalized events: A dynamic approach to linguistic relativity and determinism. In Susanne Niemeier and René Dirven (eds.), Evidence for Linguistic Relativity, 107â138. Amsterdam/Philadelphia: John Benjamins.
Slobin, Dan I. 2004. The many ways to search for a frog. In Sven Strömqvist and Ludo Verhoeven (eds.), Relating Events in Narrative: Typological and Contextual Perspectives, 219â257. Hillsdale, NJ: Lawrence Erlbaum.
Slobin, Dan I. 2006. What makes manner of motion salient?: Explorations in linguistic typology, discourse, and cognition. In Maya Hickmann and Stéphane Robert (eds.), Space in Languages: Linguistic Systems and Cognitive Categories, 59â81. Amsterdam/Philadelphia: John Benjamins.
Storch, Anne. 2014. A Grammar of Luwo: An Anthropological Approach. Amsterdam/Philadelphia: John Benjamins.
Sugiyama, Yukiko. 2005. Not all verb-framed languages are created equal: The case of Japanese. Proceedings of the Thirty-First Annual Meeting of the Berkeley Linguistics Society, 299â310. https://doi.org/10.3765/bls.v31i1.877.
Talmy, Leonard. 1985. Lexicalization patterns: Semantic structure in lexical forms. In Timothy Shopen (ed.), Language Typology and Syntactic Description, Vol. 3: Grammatical Categories and the Lexicon, 57â149. Cambridge, MA: Cambridge University Press.
Talmy, Leonard. 1991. Path to realization: A typology of event conflation. Proceedings of the 17th Annual Meeting of the Berkeley Linguistics Society, 480â519.
Talmy, Leonard. 2000. Toward a Cognitive Semantics, Vol. 2: Typology and Process in Concept Structuring. Cambridge, MA: mit Press.
Talmy, Leonard. 2017. Foreword: Past, present, and future of motion research. Motion and space across languages. In Iraide Ibarretxe-Antuñano (ed.), Motion and Space across Languages: Theory and Applications, 1â12. Amsterdam/Philadelphia: John Benjamins.
Toratani, Kiyoko. 2012. The role of sound-symbolic forms in Motion event descriptions: The case of Japanese. Review of Cognitive Linguistics 10: 90â132. https://doi.org/10.1075/rcl.10.1.03tor.
Treis, Yvonne. 2008. A Grammar of Kambaata (Ethiopia), Part 1: Phonology, Nominal Morphology and Non-verbal Predication. Cologne: Rüdiger Köppe.
Treis, Yvonne. 2023. Ideophones in Kambaata (Cushitic): Grammar, meaning and use. halOpen Science, hal-04187225v2. https://hal.science/hal-04187225/document.
Treis, Yvonne. 2024. Ideophones in Kambaata (Cushitic): Grammar, meaning and use. Linguistic Typology at the Crossroads 4, 147â200. https://doi.org/10.6092/issn.2785-0943/16393.
Van Valin, Robert D., Jr. 2005. Exploring the Syntax-semantics Interface. Cambridge, MA: Cambridge University Press.
Van Valin, Robert D., Jr. & Randy J. LaPolla. 1997. Syntax: Structure, Meaning, and Function. Cambridge: Cambridge University Press.
Wilkins, David P. 1991. The semantics, pragmatics and diachronic development of âassociated motionâ in Mparntwe Arrernte. Buffalo Papers in Linguistics 91-01: 207â57.
Talmy (2000: 25, 220, 226â227) capitalizes Motion, which covers not only events of physical motion but also events of stationariness, as well as motion components (e.g., Figure, Path). However, the present paper does not, as it specifically focuses on physical motion events.
The three vowels with a macron, Ä, Ä, and Å, appear to be the [+atr] counterparts of those without.
Kupsapiiny is characterized as predominantly anti-verb-framed because it consistently disfavors the verb-framed pattern, regardless of whether it uses a single verb or multiple verbs. The single-verb construction with a morphological satellite is clearly satellite-framed. However, the multi-verb construction in which the main verb and non-main verb express a co-event and a path of motion, respectively, is not necessarily satellite-framed, as it may not contain a satellite. See Kawachi (2025a) for further details.
Sidaama speakersâ judgment of the scope of negation in these constructions seems to be influenced by factors such as suprasegmentals and may not provide reliable data for testing the degree of syntactic integration.
Ohara (2002) and Sugiyama (2005), who found that manner-of-motion verbs in English are often rendered into Japanese mimetic adverbs, argue that speakers of a verb-framed language like Japanese do not necessarily conceptualize manner of motion as a less salient motion component.
Schaefer also suggests another hypothesis: given that Khoisan languages, which he describes as verb-framed, lack ideophones, verb-framed languages do not necessarily have ideophones, especially those in Africa. However, Schaeferâs description of Khoisan languages relies solely on Samarin (1970: 318) and Childs (1994: 197), who used secondary sources, while later studies (Killian-Hatz, 2001; Childs, 2003; Nakagawa, 2011, 2024) report that ideophones do exist in Khoisan languages.
The frequency of ideophones in descriptions of these video stimulus sets was generally not high (Matsumoto, 2025). Nevertheless, according to Matsuse (2025), speakers of Kathmandu Newar (Tibeto-Berman; Nepal), which can be analyzed as verb-framed in that one of the deictic-directional verbs always occupies the main verb position in self-agentive motion descriptions, produced 108 ideophones in 216 responses for video clips involving the manner of skipping (cf. 21 instances in 216 responses for walking clips, 7 instances in 216 responses for running clips).
Of course, the perceived unusualness of a manner of motion is subjective. Ultimately, data on all participantsâ evaluations would be necessary. Nevertheless, all the participants who watched the ssl video clips reacted by either laughing or commenting on the unusualness of the motion early in the elicitation session, unlike those who watched the medal video clips.
This contradicts Slobinâs (2004: 234) hypothesis that verb-final languages are likely to develop ideophones, which he states are preferably placed early in the sentence before the predicate to attract attention and facilitate information processing.
Ameka & Essegbey (2006: 398) also refute this hypothesis by pointing out that Ewe, a verb-serializing language, which is neither verb-framed nor satellite-framed, has ideophones.
