Constructed action

A last phenomenon that needs introduction is role shift, or constructed action (CA). This phenomenon is related to the use of space for spatial representation (Section 2.4.2.1). By shifting into the role of a character, the signer can report utterances (‘constructed dialogue’ or ‘reported speech’), thoughts, feelings and actions of that character (Metzger, 1995; see for an overview Lillo-Martin, 2012, and Cormier, Smith & Sevcikova-Sehyr, 2015 ).

Of course, constructed dialogue/reported speech exists in both signed and spoken languages. Yet, in sign languages, constructed dialogue often (but not always) concurs with descriptions of actions, emotions and behaviors of the portrayed character (Metzger, 1995). The latter ‘non-quotative uses’ are observed in speakers as well (Cormier, Smith & Zwets, 2013), albeit to a lesser extent. During CA sequences, the signer uses the real-life-sized viewer space or character perspective (Section 2.4.2.1, Figure 2.8). The signer can use one or more of the following strategies to portray a character:

• Use of pantomime to demonstrate how a character moves or acts, including the use of the hands to show how a character handles or manipulates objects (i.e., Handle classifiers, Figure 2.5b);

• Use of constructed dialogue to report what is being signed;

• Adaption of signing style to mimic the signing style of the character, or to distinguish between two or more characters;

• Use of facial expressions to display the character’s facial expressions (e.g., emotions);

• Use of the torso to mimic the torso movement or posture of a character;

• Use of head turn or eye gaze that mimic those of the character.

There are different ways in which a signer can signal that he or she is taking on the role of a character. There seems to be general agreement that gazing away from the addressee (thus, breaking eye contact) is a crucial and obligatory marker to indicate role shift (e.g., Loew, 1984; Padden, 1986;

Herrmann & Steinbach, 2012; Cormier et al., 2015). Other, optional, signals are changes in body position and facial expression.

Engberg-Pedersen (1993) discusses three phenomena related to constructed action (‘role shift’ in her terms): (i) shifted attribution of expressive elements, (ii) shifted reference, and (iii) shifted locus. Shifted attribution of expressive elements indicates that the feelings and attitudes expressed on the signer’s face must be attributed to the character, and not to the signer. Shifted reference refers to the given that, during CA sequences, pronouns should be interpreted from the character’s point of view. The moment a signer takes up a role, the portrayed character is mapped onto the signer’s body, and consequently, a point to the chest (IX1) should be interpreted as referring to the character rather than to first person. Shifted locus refers to the position of the portrayed character and other characters in the quoted context. The eye gaze of the signer should be interpreted as the eye gaze of the character, directed at the location(s) of the other character(s). This is exemplified in Figure 2.12. Suppose that the scene involving a woman riding a bike and passing two children sitting on a fence (girl on the left, boy on the right) is portrayed by the signer from the viewpoint of the girl. A gaze directed towards the front-right area of the signing space should then be interpreted as a gaze towards the (imaginary) woman, while a gaze directed to the left of the signer should be interpreted as a gaze towards the boy. Likewise, agreement verbs and orientations of the head and body towards loci should be interpreted from the point of view of the character.

Figure 2.12. Examples of eye gaze of a character portrayed by the signer, i.e., under role shift, directed to the locations of the (conceptualized) other characters. (photo:

©Eveline Boers-Visker).

If the signer took up the role of another referent, say, the woman in Figure 2.12, a rotation of the referent space would occur and consequently, the movement, facial expressions, eye gaze, pronouns, and agreement verbs would have to be interpreted as belonging to or referring to this new character. When describing an event with multiple characters, the signer must switch back and forth between different perspectives, and signal these switches clearly to the conversation partner. In between, the signer can

‘return’ to the narrator-role (observer perspective) to comment on the situation.

In the example above, the signer takes on the role of characters and the narrator in turn. However, it is also possible to represent two or even three referents simultaneously, using different articulators, i.e., the hands and face (‘body partioning’, Dudis, 2004). This is exemplified in Figure 2.13, showing a still from an NGT signer narrating the Frog Story.³⁴

Figure 2.13. Example of signer simultaneously representing two entities on different articulators (stills from Corpus NGT, used under Creative Common license cc by-nc-sa 3.0 nl).

In stills a and b, the body, head, face, and one of the hands depict one referent, while a second referent is depicted by an Entity classifier predicate on the other hand. In still b, one character is mapped on the body, including

34 The stills are taken from clip 403 from the Corpus NGT (www.ru.nl/corpusngt). The Frog Story (Frog where are you?, Mercer Mayer, 1969) is a wordless picture story often used for narrative elicitation.

facial expression, and the other characters (namely, bees) are represented by means of Entity classifier predicates.

Clearly, the use of CA constructions involving multiple perspectives and/or the simultaneous use of different articulators to represent different referents adds an extra layer of complexity, requiring both cognitive and linguistic skills that can be challenging for SL2-learners (McKee & McKee, 1992).

In sum: SL2-learners have to recognize (i) whether a specific type of information is best presented by a spatial or an abstract representation, (ii) which spatial devices should be recruited to express the information, and (iii) when a mental rotation of the space is required, and how it should be performed. For each type of device, learners additionally have to acquire the rules and constraints that govern that device. In case of constructed action (a form of spatial representation), an extra layer of complexity is added, as the locations of referents in sign space are different, according to the point of view of the various characters and the narrator. This has consequences for reference to entities in space, and for the use of pronouns and agreement verbs.