Acquisition and emergence of Entity classifier predicates

Figure 2.14. Z-axis, x-axis and diagonal axis.

Emerging sign languages give insight into the strategies people use to indicate the verb’s arguments when the sign language has not (yet) developed an agreement system. Two strategies that have been observed in emerging sign languages are (i) successive expression of the subject and the object in separate clauses (‘successive 1-argument structures’, e.g., MAN GIVE,

WOMAN RECEIVE, see Chapter 5) (Senghas, Coppola, Newport & Supalla, 1997 for ISN; Meir, 2010 for ABSL and ISL; Ergin et al., 2018 for CTSL) and (ii) a strategy termed ‘character assignment’ by Ergin et al., in which the signer uses his or her own body as stand-in for the subject, and sometimes the adressee’s body as stand-in for the object (e.g., “I am the man, you are the woman, I give you the book”; see Chapter 5) (Ergin et al., 2018; Meir, 2010).

Meir (2010) describes a third strategy, namely an instance of use of an auxiliary-like element similar to the one identified in the L1 data by Morgan et al. (2006) (see above). Meir (2010) considers this element as a form “that can be regarded as precursor[s] of verb agreement” (p. 118). We found multiple instances of these strategies, as well as a preference to use the z-axis, in the SL2-data we obtained (see Chapter 5).

2.4.7.3 Acquisition and emergence of Entity classifier predicates

There is a growing body of studies that analyze L1 production data on Entity classifier predicates (henceforth: classifier predicates). The picture that emerges from these studies is that children are able to use classifier predicates in appropriate contexts with moving or static objects at a young age (Kantor, 1980; Schick, 2006; Slobin et al., 2003). However, their production is prone to errors, and it takes several years, up to 9;0, to master

the system completely (Newport & Meier, 1985; Hoiting & Slobin, 2002;

Baker et al., 2008). Reported errors are:

• Substitution of the classifier handshape (Supalla, 1982; De Beuzeville, 2006);

• Omission of components (e.g., manner of movement; Newport &

Meier, 1985);

• Sequential production of complex movement patterns (e.g., a straight upward movement followed by an arc instead of an upward arc movement; Newport & Supalla, 1980);

• Failure to introduce referents (Slobin et al., 2003; Tang, Sze & Lam, 2007);

• Omission of the Ground object (Supalla, 1982; Newport & Meier, 1985; Slobin et al., 2003; Engberg-Pedersen, 2003; Tang et al., 2007;

Sümer, 2015);⁴⁶

• Failure to produce Figure and Ground simultaneously, instead expressing both objects sequentially (Supalla, 1982; Tang et al., 2007);⁴⁷

• Signing outside the signing space (De Beuzeville, 2006).

Morgan (2002), De Beuzeville (2006), Tang et al. (2007), and Tang and Li (2018) report children employing avoidance strategies such as production of lexical descriptions instead of classifier predicates, role shift, or use of the whole body as stand-in for an animate referent (‘whole-body language’).

Kantor (1980) notes that children sometimes employ classifiers in simple contexts, but avoid to use classifiers for similar entities in complex environments.

There are only few publications on the use of classifiers in SL2-learners.

Marshall and Morgan (2015) elicited spatial representations in novel learners of BSL. Ferrara and Nilsson (2017) investigated the use of classifiers by novel learners of Norwegian Sign Language (NSL) in longer stretches of text.

Frederiksen and Mayberry (2019) investigated the frequency of referential

46 The notions of Figure and Ground have not been introduced yet. In two-handed classifier constructions, the Ground object is the stationary, and usually the bigger object. The Figure object is moving (or could move) in relation to the Ground object.

See Section 4.2.1.3.

47 But see Footnote 7 in Section 4.2.3.

devices, including classifier predicates, in elicited short stories, signed by novel signers. Typical errors reported in these studies are:

• Difficulties in choosing the correct classifier handshape (Marshall &

Morgan, 2015);

• Difficulties in coordinating both hands in two-handed constructions (Ferrara & Nilsson, 2017);

• Difficulties in planning the scene in relation to the body (Ferrara &

Nilsson, 2017);

• (Unappropriate) substitution of classifier entities with signs marked for location (Ferrara & Nilsson, 2017);

• Overuse of classifier predicates to reintroduce referents in a narrative by learners, in contexts where L1-signers do not use classifier predicates (Frederiksen & Mayberry, 2019).

Ferrara and Nilsson (2017) report avoidance strategies similar to those found by Morgan (2002) in L1-learners, that is, resorting to the production of lexical signs instead of using classifier predicates.

Given the fact that non-signers are found to use handshapes to represent referents in gestures (Section 2.4.6), it is not surprising that these forms have been found in early stages of emerging sign languages. Goldin-Meadow, Brentari, Coppola, Horton and Senghas (2015) show that users of the emerging ISN and Nicaraguan homesigners both use the hand to represent objects. The set of handshapes used for these representations, however, is more consistent in the ISN-signers than in the homesigners, which suggests that conventionalization has taken place. Aronoff et al. (2003) elicited classifier predicates in ISL-signers, and compared these with the productions of ASL-signers. This comparison is interesting given the respective age of both languages: ISL is a young language, whereas ASL is among the oldest sign languages known. Two interesting results emerged from the analyses. First, the ISL-signers frequently produce what the authors call ‘referent projections’, that is, they use the body to enact the referent (e.g., move like a cat). In contrast, ASL-signers prefer to use Entity classifier predicates instead. Secondly, a comparison of the classifier predicates used in both languages reveals that ASL classifier predicates have developed to be less iconic and more abstract than ISL classifier predicates. ASL, for example, has two Entity classifiers that have a broad class membership: one to denote a class of ‘vehicles’ and another to denote a class of ‘objects’. ISL, in contrast,

does employ a classifier for vehicles, but this class is much smaller than the class of vehicles the ASL vehicle-classifier refers to. The authors attribute these differences to language age.