Chapter 2 Literature Review
2.4. Classifier research in sign languages
2.4.2. Classifiers a morphosyntactic subsystem
The term morphosyntactic subsystem refers to morphological elements, which are
relevant for syntactic purposes. The examples of the different types of classifiers in
spoken languages are illustrative of this term, e.g. the use of numeral classifiers in
Japanese. This means these constructions form part of the grammar of Japanese and
cannot be left out of a sentence.
Schembri agrees that handshape units in sign language classifiers “form part of a
morphosyntactic system in signed languages” in relation to the lexicon (2003: 20). He
acknowledges that applying this criterion to sign languages is not clear cut. He claims
that class terms in spoken languages (Grinevald, 2000) is a more applicable category for
a comparison with classifier handshapes.
Grinevald explains class terms as follows: “[They] are lexical in the sense that they
operate on derivational or compounding morphology at word level" (2000: 58). They do
not have the same function of spoken language classifiers as described earlier, but they
/ lexicalised / grammaticised classifier handshapes in comparison to class term
morphemes in English such as ‘-berry’ and derivational morphemes such as ‘–er’
(manager) and ‘-ism’ (skepticism). In terms of sign languages, some classifiers become
part of the lexicon (i.e. they are no longer creative signs which are context-bound). In
this way they are similar to class terms.
The examples below in Figures 2-3 (a)-(c) are SASL examples of frozen classifiers:
MEETING (from the category entity classifiers), KEY / LOCK (from the category handling
classifiers) and RECEIPT (from the category SASS). Schembri points out that although the
class term comparison is more valid, classifiers in spoken language do not play a major
role in lexicogenesis. This conclusion seems to indicate that Grinevald’s
‘morphosyntactic subsystem’ is not really applicable to sign language classifiers.
a) MEETING b) KEY c) RECEIPT
In light of the above, the criterion of ‘morphosyntactic subsystem’ seems to be
irrelevant for sign languages. However, Schembri offers the potential solution by
pointing out that sign languages may display a unique morphosyntactic subsystem which
has no equal in spoken languages, “… unique in its fusion of linguistic and visuospatial
properties” (2003: 21).
Schwager and Zeshan state that “the morphology of (these) sign languages is largely
simultaneous rather than sequential, and involves modulations in signing space typical
of the sign language modality” (2008: 535). Although they do not discuss classifiers in
particular, they tabled a number of morphological criteria observed in three sign
languages they studied.
In this table, they identify “class agreement” as a grammatical category under “event
concept class” and the morphological process related to this as affixation. They explain
this as follows: “… the handshape representing the referent class, such as the human
and vehicle referents … is a bound morpheme that is combined with morphemes of
location and movement in an affixation process, whereby a complex multimorphemic
sign is created” (2008: 538-539).
The use of classifiers in sign languages does not constitute such a highly grammaticalised
system as, for example, in Japanese. They behave differently from spoken language
classifiers. However, in the context of sign languages, they do occur on a highly frequent
2.4.3. Classifiers - semantically motivated systems of classification that do
not classify all nouns
In relation to the third criterion, Schembri raises the question of what exactly is
classified by handshape units in sign languages. According to him, the general
understanding of spoken language classifiers is that “a classifier classifies in the sense
that it denotes some salient or perceived characteristic of the referent represented by
an associated noun” (2003: 21). This means that they are semantically motivated. This
seems to be the basis for part of Sutton-Spence and Woll’s criteria for sign language
classifiers mentioned earlier (1999), as well as Valli and Lucas’ definition (2000).
It is possible that the use of classifiers in sign languages does not actually fully conform
to the above characteristics. Engberg-Pedersen (1993: 246), uses the term
polymorphemic verbs for classifiers in sign languages and argues that although
polymorphemic verbs in Danish Sign Language distinguish between animate and
inanimate, the “sheer number of handshape units that can be used in polymorphemic
verbs of human beings is evidence that the handshape units have more than
classificatory meaning.” She also claims that the qualities of the referent are not the
only factors that determine the choice of handshape. Movement and handshape are
mutually interdependent, and for this reason, the argument that handshape units are
classifiers cannot be justified.
It would seem that not all sign language linguists agree on the point of classifiers
representing salient characteristics. For example, Aronoff et al. state that “entity
upright human, seated human, vehicle, and so on, also enter into complex constructions,
by combining with other classifiers (signed by the other hand) as well as with different
movement roots indicating path shapes and manners of movement” (2005: 23).
Zwitserlood et al. offer the following explanation of classifiers: “Classifier predicates are
morphologically complex predicates in which the classifier refers to a referent.
Referents are classified on the basis of salient characteristics like shape (classifiers for
long and thin entities, flat entities, and round and/or cylindrical entities are found in
almost all sign languages), animacy (some sign languages have special classifiers for
humans and animals or for men and women) or other characteristics” (2012: 1639).