Qualitative analysis

In the previous section, we quantitatively described the learning process of the SL2-participants regarding use of space. The following is a brief synopsis of the extended logbook, in which we noted remarkable L2-behavior regarding use of space.

3.4.2.1 Pointing signs

When we considered the pointing behavior of both SL2-participants, we noticed a number of instances of stacking during the first and second years, that is, using the same location in space for more than one referent (Loew, 1984). Loew found this behavior, illustrated in Figure 3.14, in children acquiring sign language as an L1.

Figure 3.14. Example of stacking.

Another striking observation is the misuse of pronouns during constructed action. In some cases, but not always, Anna struggled with using the correct pronoun when reporting speech or actions from other persons.

An example, judged as non-target-like by our deaf informants,¹⁰ is shown in Figure 3.15, in which Anna intended to sign ‘you’ but instead pointed to an abstract location in signing space. We observed this behavior until the beginning of year 4.

10 One of the reviewers raised the question that an alternative interpretation (e.g.,

“he informed me that I missed him”) could be possible too. However, the Dutch mouthings that accompany the pointing signs point to one possible interpretation (“I miss you”).

Figure 3.15. Misuse of pronouns during constructed action.

3.4.2.2 Verbs

We found that the SL2-participants started to modify non-iconically motivated agreement verbs intentionally at the end of year 1. In their first year, they produced many verbs that would have been candidates for agreement in their citation form; however, this decreased in the following years. We noted a few reversal errors (e.g., 3OBSERVE1 instead of 1OBSERVE3), but the majority of the agreement verbs in the observer perspective were produced correctly. In character perspective, we saw Anna struggle to produce the correct inflections after a shift of reference.

We did not observe cases of overgeneralization, that is, instances of plain verbs that were spatially modified.

As noted in the previous section, both L1- and SL2-participants rarely used classifier predicates. Initially, the SL2-participants did not produce classifier predicates at all. We noted the first use of classifier predicates in their data after eight months (Anna) and 18 months (Charlotte). It is noteworthy that the first classifier predicates to emerge in Anna’s data were restricted to the -classifier for a moving person and the -classifier (palm facing down) to represent a crowd of people. The first instances of the -classifier occurred several times, but Anna did does not vary the location and movement; she seemed to encode a prototypical event schema (viz. “the person approached me”). Charlotte started to produce classifier predicates much later and, like Anna, she initially produced the -classifier to represent a crowd and the -classifier to depict standing and walking persons. However, in contrast to Anna, Charlotte immediately produced the

-classifier in different, less prototypical, variations. She produced the -classifier as well, to indicate the location of several podia. During years 3 and 4, both SL2-participants produced a variety of classifier predicates. However, Charlotte had issues with selecting the appropriate classifier handhapes, as well as with producing the correct phonological form. The stills in Figure 3.16a show an example of the selection of an inappropriate classifier handshape. Instead of depicting a row of chairs, she selected the classifier for a row of standing people. In addition, she struggled with phonology, producing three rows of stairs with a slightly downward movement, which should be upwards. The classifier predicates denoting sitting persons in Figure 3.16b had an incorrect orientation (viz. legs up in the air). Both examples were rejected by our deaf informants. Charlotte struggled with the selection of appropriate classifier handshapes and the spatial configuration of classifier predicates until the beginning of year 4.¹¹

Figure 3.16. Examples of L2-participant struggling with handshape selection and spatial configuration of Whole Entity classifier predicates.

Both SL2-participants used classifier predicates to describe the actual layout of a scene (for example, to provide a description of a classroom or an interpreter setting), as well as classifier predicates that should be interpreted

11 At the end of year 3 (session 3D), six of the seven classifier predicates were either inappropriate, phonologically incorrect, or lacked a referent. In the last session (session 4B, after 3.5 years of instruction), still two out of eight classifier predicates were off-target.

metaphorically (Sandler & Lillo-Martin, 2006). In the latter case, the movement or location of the classifier predicate was not meant to encode the real physical activity, but beared a metaphorical meaning. The location of the referent denoted by the right hand in the construction, which is translated as “he supported me” in Figure 3.17, for example, was (as the context revealed) not literally behind the signer.¹² We noted this non-literal, metaphorical use during the sessions recorded in years 3 and 4. Once metaphorical classifiers appeared in the data, both participants showed creative use of the constructions, that is, they modulated handshapes according to their needs and added repetitions or multiple movements (e.g.,

“join different groups” or “two people taught the group”).¹³

Figure 3.17. Examples of metaphorical use of classifier predicates.

12 The construction is related to the Dutch idiomatic expression “he stands behind me,” meaning “he supports me when I need support.”

13 One could argue that the constructions are lexicalized forms that are de-lexicalized during the discourse. However, we do not have evidence that the forms displayed by the students are lexicalized forms in NGT.

3.4.2.3 Signs marked for location

Whereas the classifier predicates posed difficulties for the SL2-participants, they seemed to acquire signs marked for location quickly and without much difficulty. From the middle of the second year, Anna started to utilize the possibility to create a compound¹⁴ by attaching the sign PERSON or AREA to a noun (for example TEACHER^PERSON). The fact that Anna marked the second part of the compound (and sometimes the first part, too) for location and used this strategy to localize signs that can exist without the added segment (i.e., the NGT signs TEACHER, DENTIST, or WITCH can exist without the added segment PERSON) signals intentional use of this strategy. Charlotte started to use this strategy in year 3, although she produced one example as early as session 1B.

While we did not find examples of overgeneralization in agreement verbs, we did notice overgeneralization in spatially modified nouns and adjectives. Figure 3.18 shows an example: Anna produced the sign

SINTERKLAAS (i.e., a Dutch mythological figure like Santa Claus), a compound consisting of the sign BEARD followed by the sign SCEPTER, which is produced next to the signer’s body and cannot be spatially modified. However, Anna detached the second part SCEPTER from the original location and produced it at the location where the referent was situated.

Figure 3.18. Example of overgeneralization of spatial modification of a (compound) noun.

14 Some scholars consider these examples as derivation instead of compounds (e.g., PERSON is considered an agentive suffix) (Sandler & Lillo-Martin, 2006). However, such an analysis is not applicable to NGT, as the compound members in these examples do appear as free forms in the language, while affixes do not appear as a free form (Meir, 2012b).