Background. Many of the world’s spoken and sign languages mark the difference between the two major lexical categories noun and verb (Rijkhoff, 2007; Haspelmath, 2023). In the case of German Sign Language (DGS), different morphophonological properties for nouns and verbs have previously been described (e.g. Happ, 2005) but it remains unclear to what extent these strategies for overtly marking word classes apply systematically or maybe even generalise across the distinction of single noun-verb pairs. It has been observed that sign languages make use of gestural demonstrations in different grammatical domains (Aronoff et al., 2005). Promi-nent examples are the expression of telicity (Wilbur, 2003; Strickland et al., 2015), agreement (Schembri et al., 2018; Pfau et al., 2018), classifiers (Schembri et al., 2005; Zwitserlood, 2012), and role shift or constructed action (CA) (Davidson, 2015; Steinbach, 2023). Therefore, one particularly interesting question in this context is whether sign languages also use different iconic strategies such as e.g. gestural demonstrations to distinguish between nouns and verbs (Padden et al., 2015; Abner et al., 2019).
In two consecutive experiments, we investigate morphophonological markings distinguishing nouns and verbs in DGS with a special focus on the form and function of gestural demonstra-tions: Experiment 1 employs an elicitation paradigm in which signers watched videos depicting scenes specifically designed to elicit either the nominal or the verbal use of a potentially multi-functional DGS sign for concrete objects or actions in the participants’ responses. Experiment 2 uses a sentence reproduction task (Pendzich et al., 2022) designed to elicit either the nominal or verbal use of concrete as well as abstract DGS signs within a given syntactic context. At the time of writing, data analysis for Experiment 1 and data collection for Experiment 2 are still ongoing, which is why we only present preliminary findings from the former here.
Experiment 1. Deaf signers (N = 9, mean age = 40.8) watched videos (N = 91) depicting scenes specifically designed to elicit either the nominal or the verbal use of a potentially multi-functional DGS sign. Uses of the target signs were annotated in ELAN and classified as either nominal or verbal by a coda, whereas unclear cases were evaluated by a deaf member of the research team who is a native signer of DGS. In total, we collected 747 signed sequences which will be evaluated against the following criteria: (1) mouthing vs. mouth gesture, (2) head and body movement, (3) duration of produced signs, (4) reduplication and usage of sign space, (5) interaction of lexical signs with CA, and (6) sequential supplement for nominal signs in the form of size-and-shape specifiers (SASS). Because data analysis is still ongoing, we here pre-sent first results limited to four representative noun-verbs pairs (PLAY, SLEEP, ICE-CREAM, and HAIRDRY) and the evaluation of the two criteria (1) and (5), which both involve gestural demonstrations. This suggests a largely systematic morphophonologial marking of nominal and verbal uses with regard to mouth gesture and mouthing (Figure 1, top panel). Whereas mouth-ings co-occur with both nominal and verbal uses of a sign, mouth gestures are primarily pro-duced as part of verbs. The iconic potential of a sign’s phonological form seems to impact the frequency in which mouth gestures are produced alongside verbs and mouthings alongside nouns (cf. ICE-CREAM and HAIRDRY). Iconic verbal signs such as HAIRDRY can be context-dependently supplemented by CA involving a gestural demonstration of the action described by the verbal sign (Figure 2, bottom panel). Significantly, elements of CA seem to primarily occur simultaneously as modifications and/or supplements to a verb’s phonological form, while pho-nologically and semantically underspecified nouns tend to be complemented by sequentially occurring SASS.
Experiment 2. We will employ a sentence reproduction task (Pendzich et al., 2022) specifically designed to elicit either the nominal or verbal use of concrete as well as abstract DGS signs within a given syntactic context by deaf signers participating in the experiment. In addition to video recordings and manual annotation we will also collect data about non-manual elements during sentence reproduction using the so-called TrueDepth camera of an iPhone. We will combine this automated capturing of non-manual components with the analysis of facial ex-pressions using the Facial Action Coding System (FACS) by Ekman et al. (2002), which will enable us to adequately quantify non-manual components accompanying nouns and verbs.
Discussion. The preliminary analysis of data from Experiment 1 supports the notion that DGS exhibits some systematic morphophonological marking of nominal and verbal uses of a sign. Signs with low iconic potential (e.g., PLAY) seem to be accompanied by mouthings regardless of whether the sign is used in a verbal or nominal context. In contrast, for signs with higher iconic potential, mouthings and mouth gestures seem to more clearly differentiate between nom-inal and verbal uses. The latter strategy, i.e. mouth gestures, involves a non-manual gestural demonstration of the corresponding verbal activity. In addition, with iconic signs, signers use more frequently CA to distinguish between the verbal and the nominal use of a sign. Again, this strategy involves a gestural demonstration of the verbal activity. Our data suggest that the iconic potential of a sign’s phonological form and the respective context mediate whether a strictly conventionalised lexical strategy or a combination with CA may be used. Signs with a high iconic potential are more likely accompanied by a gestural demonstration in the verbal use.
Singers can thus flexibly combine lexical as well as manual and non-manual gestural strategies for marking word class distinctions by modulating the iconic potential of a signs’ phonological form with gestural demonstrations. Gestural demonstrations are thus one strategy used in DGS which is systematically exploited to distinguish verbs from nouns. The iconic component of signs that depict, for instance, the handling of an object can be modified by adding a more elab-orated gestural demonstration of the activity depicted. This demonstration typically involves a manual as well as a non-manual component that is typically added in verbal contexts, i.e. to depict an action. For the depiction of objects, other strategies such as e.g. classifiers and mouth-ing are used.
Following Davidson (2015 and Henderson (2016), iconic signs such as HAIRDRY in (1) can be analyzed as mixed items that involve a conventionalized meaning component and a gestural demonstration (marked in bold). In verbal contexts, this second meaning component can be actively modulated to express various aspects of the event demonstrated. In contrast, the lexical entry of non-iconic signs such as PLAY in (2) does not involve a gestural demonstration. Therefore, a modulation of this component is not possible.
In summary, manual (CA) and non-manual (mouth gesture) gestural demonstrations are an indirect indicator for a verbal use of a sign and thus an iconic strategy that can be used to dis-tinguish verbs from nouns. We are confident that a full analysis of the results of Experiment 1 and Experiment 2 will allow us to characterise the manual and non-manual iconic components marking nouns and verbs in DGS and the relevance of gestural demonstrations for word class distinctions.