Spatial modification of verbs

A second mechanism that uses space to create meaning is found in the verbal system. Across sign languages, a subset of verbs can be spatially modified such that they move between locations in space and/or are oriented towards a location or are produced at a specific location. Such modifications enable the signer to provide information about who is doing what to whom, in what direction a referent is moving or is moved to, or where it is located. This phenomenon has been identified in a wide variety of sign languages (for overviews, see Lillo-Martin & Meier, 2011 and Mathur & Rathmann, 2012), including NGT (Bos, 1994, 1995; Zwitserlood & Van Gijn, 2006).

We will explicate spatial verb modification on the basis of the highly influential tri-partite classification system originally proposed by Padden (1988) for ASL. This classification, which has been widely adopted by other researchers and for other sign languages, is based on the observation that verbs can or cannot combine with different kinds of spatial morphemes. It must be noted, however, that not all scholars agree upon this morphological analysis (e.g., Johnston, 1991). Padden (1988) distinguishes three verb classes: agreement verbs,³ spatial verbs, and plain verbs. Agreement verbs provide information about the subject and/or the object of the verb by modifying the movement trajectory and/or the orientation of the hand(s).

Figure 3.5 depicts inflected forms of the NGT verbs GIVE and CALL-SOMEONE.

GIVE involves a movement trajectory: the initial position of the hand corresponds to the locus associated with the subject of the verb, whereas the final position of the hands aligns with the locus of the indirect object argument. CALL-SOMEONE agrees with its direct object by orienting the palm of the hand towards the object of the verb (Meir, 2002). Other verbs (for

3 Initially, Padden (1988) used the term inflecting verbs. Later, she adopted Johnston and Liddell’s (1987) proposal to rename this category agreement verbs, as plain verbs can also display inflection for aspect (Padden, 1990).

example, SEND) combine a movement trajectory and a change in orientation.

One can distinguish regular verbs, which move from subject to object, and backward verbs, in which the direction of movement is reversed (for example

FETCH).

Figure 3.5. Examples of agreement verbs (still from signed story ‘Haas wil worteltjestaart’, Kentalis Multimedia Haren).

Not all agreement verbs can be marked for both the subject and the object; some verbs can only be inflected for object (‘single agreement,’ for example, COMPLIMENT in Figure 3.5). This lack of marking can be due to phonological constraints, for instance, when the verb is body-anchored, as is the case for COMPLIMENT (Rathmann & Mathur, 2003). Even if a verb is candidate for ‘double agreement’ (i.e., the verb can in principle be inflected for both subject and object), signers do not always realize agreement. They can opt to use the citation form (thus not realizing agreement at all) or mark the verb for subject or object only. In the latter case, a verb is partially modified (Mathur & Rathmann, 2010). A subset of agreement verbs can embed a Handle classifier: a handshape that denotes how an object is handled or manipulated. GIVE in Figure 3.5, for example, is produced with a flat, closed handshape, signaling a flat object (in this context, a piece of paper).

Spatial verbs can also be modified, but this modification does not signal agreement with subject or object. Instead, spatial verbs move between locations, signaling the begin location and/or the end location of a movement. The spatial verb PUT (Figure 3.6), for example, moves between the location where the box was (source) to the location to which the box is moved (goal). In our analysis, we followed Padden (1988) in distinguishing between locative verbs and classifier verbs (Whole Entity classifier predicates). Locative verbs combine with locative morphemes (GO and GRAB

in Figure 3.6), and some of them also contain a Handle classifier handshape (PUT in Figure 3.6).

Classifier predicates contain a so-called Entity classifier. Like locative verbs, they signal information about the source and/or goal of the movement by combining with locative morphemes. In addition, the Entity classifier denotes information about the nominal, reflecting characteristics of the referent. We will discuss classifier predicates separately later in this chapter.

Figure 3.6. Examples of spatial verbs (stills from signed story ‘Haas wil worteltjestaart’, Kentalis Multimedia Haren).

In contrast to agreeing and spatial verbs, plain verbs (for example, the NGT signs EAT, LOVE, and WANT) cannot mark their subject or object referents or begin and end locations by altering their movement trajectory or orientation, nor do they combine with a Handle or Entity classifier. Plain verbs can inflect for manner (e.g., “to eat fast” or “to eat a lot”). However, some plain verbs can be produced in a non-neutral location. The sign BE

-PRESENT in Figure 3.7, for example, is produced at a non-neutral, specific location, namely the location of the person the verb agrees with. Padden (1990) argues that these verbs do not take agreement morphology.

Zwitserlood and Van Gijn (2006) and Costello (2016) disagree; they classify these cases as examples of spatial agreement. In our analysis, we treated

these particular ‘localized plain verbs’ as instances of agreement verbs, using Costello’s (2016) label ‘single argument agreement’. Not all plain verbs are candidates for de-localization. Body-anchored signs, such as EAT and LOVE, cannot undergo this type of spatial modification (Padden 1988).

Figure 3.7. Example of single argument agreement verb BE-PRESENT (stills from http://www.gebareninzicht.nl/).