ESPECIAL Línguas Gestuais

Transcription

Cadernos
de Saúde
NÚMERO
E S P E C I A L
Línguas Gestuais
VOLUME 2, 2009
Publicação Semestral
Cadernos de Saúde  Vol u m e 2  Número especial de Linguagem Gestual
Índice
Editorial
5
Alexandre Castro Caldas
Ana Mineiro
Brain Mechanisms for Sign Language
7
Phonological category resolution:
A study of handshapes in younger and older sign languages
13
Categorias fonológicas:
um estudo sobre orientação manual em línguas gestuais antigas e novas
Assaf Israel and Wendy Sandler
Imperativos Análogos a Raízes Infinitivas:
Evidência das Línguas de Sinais Americana e Brasileira
29
Imperatives function as analogues to root infinitives:
evidence from ASL and LBS
Diane Lillo-Martin e Ronice Müller de Quadros
Sign production by first-time hearing signers:
A closer look at handshape accuracy
Produção de gestos por ouvintes gestuantes iniciados:
um olhar atento para a orientação manual
Deborah Chen Pichler
37
Oral language and sign language:
possible approaches for deaf people’s language development
51
Lingua orale e lingua dei segni:
approcci possibili per lo sviluppo del linguaggio nei sordi
Carmela Bertone and Francesca Volpato
LÁBIOS, LEITE, CHOCOLATE, LARANJA, ETC.:
um estudo sobre os nomes das cores em LSB
63
LIPS, MILK, CHOCOLATE, ORANGE, ETC.:
a study of colors’ names in Brazilian Sign Language
Sandra Patrícia de Faria do Nascimento
Adding pieces to the Portuguese Sign Language lexicon puzzle:
three pilot studies
83
Juntando mais peças ao puzzle do léxico da LPG:
três estudos pilotos
Ana Mineiro; Joana Pereira; Liliana Duarte and Isabel Morais
Cognitive Studies on Portuguese Sign Language (LGP):
a work in progress
Estudos Cognitivos em Língua Gestual Portuguesa:
estudo de arte
Maria Vânia Silva Nunes; Paulo Vaz de Carvalho; Ana Mineiro and Alexandre Castro Caldas
99
Editorial
Ana Mineiro
O Instituto de Ciências da Saúde (ICS) foi
vise a formação de docentes de Língua Gestual
criado na Universidade Católica Portuguesa
Portuguesa, nomeadamente surdos, falantes
em 2004. Foi intenção da Universidade orga-
fluentes de LGP, proporcionando formação
nizar este domínio do saber, de forma ampla
científica nesta língua falada com as mãos,
e integrando várias áreas do conhecimento,
à semelhança daquilo que ocorre em países
entre as quais, o ensino e investigação em
como os Estados Unidos e o Brasil.
Língua Gestual Portuguesa.
Tendo em mente este objectivo e consolidada
Em 2006, foi lançado pelo ICS o primeiro
a tradição do ICS na área da Língua Gestual
de Língua Gestual Portuguesa e Educação de
Portuguesa, a UCP irá lançar, em 2009, uma
Surdos, através de um Protocolo com a Casa
Licenciatura em regime de Ensino a Distância,
Pia de Lisboa e com a inestimável colaboração
leccionada em Língua Gestual Portuguesa.
e o apoio constante dos Senhores Doutores
O início desta Licenciatura estará sempre
Maria Augusta Amaral e Amândio Coutinho.
associado à investigação. O facto de a LGP
Esse Mestrado teve a particularidade de ser o
ser uma língua pouco investigada e não ter
primeiro no país, repetindo-se, numa segunda
sido alvo de atenção sustentada por parte
edição em 2008.
dos linguistas portugueses, levou a UCP a
A Língua Gestual Portuguesa é reconhecida
sedimentar o plano de estudos da Licenciatura
pela Constituição (Lei Constitucional nº 1/97,
e a concepção da mesma numa estreita ligação
artº 74º, alínea h) como a língua através da qual
com a investigação nesta língua. A Fundação
se deve desenrolar o ensino-aprendizagem das
Calouste Gulbenkian apostou neste projecto e
crianças surdas portuguesas. Neste sentido, é
financiou a dimensão do ensino-investigação
necessário criar um quadro universitário que
desta Licenciatura, através do Projecto «Turma
6
Cade rn o s d e S a ú d e  Vo l. 2  Nú m e ro es p ec i a l d e L í n g ua s G es tua i s
Piloto de LGP». A Fundação PT cedeu a pla-
da Universidade Federal de Santa Catarina,
taforma para o Ensino a Distância apoiando
Deborah Chen Pichler, da Universidade de
este projecto pioneiro. Outros projectos que
Gallaudet, Diane Lillo Martin, da Universidade
envolvem a forma como os Surdos pensam,
de Connecticut, Sandra de Faria, da Universi-
aprendem e falam encontram-se em fase de
dade de Brasília, Carmela Bertone e Francesca
desenvolvimento.
Volpato, da Universidade de Veneza.
Foi nossa preocupação promover a Conferên-
Este número especial dos Cadernos de Saúde
cia Internacional, Sign Languages Around the
reúne assim os artigos dos investigadores que
World de forma a partilharmos e a discutirmos
connosco partilharam os três dias que durou
os aspectos científicos das línguas gestuais com
a conferência.
alguns dos investigadores mais conceituados no
Agradecemos o apoio pecuniário da Funda-
mundo, nomeadamente com Wendy Sandler, da
ção para a Ciência e a Tecnologia, da Fundação
Universidade de Haifa, Ronice Muller Quadros,
Calouste Gulbenkian e da Casa Pia de Lisboa.
Alexandre Castro Caldas1*
1
Instituto de Ciências da Saúde – Universidade Católica Portuguesa
General Framework
The use of modern technologies to explore brain
function stimulated the discussion of the proper
nature of cognition and behavior. We now have
to admit that individual cognitive competences are
the result of a convergence of several components.
Some of these components are related to biology,
some are related to environmental stimulation and
some are related to the opportunity.
Language is one the competences that made
human kind different from the other species. We
hardly understand how this phenotype developed
with the development of humans and it is necessary
to work on hypothesis based on several kinds of
findings.
If we take in consideration the average situation
of a middle aged Caucasian human being that is not
left hander that acquired oral language in a normal
speaking monolingual community we can discuss
the cortical organization of speech processing like,
for instance, Hickok and Poeppel did (2007). These
authors considered that the cortical areas involved
in speech processing are organized in two main
pathways: a dorsal stream that maps acoustic speech
signals to frontal lobe articulatory networks and a
ventral stream that processes speech signals for
comprehension. This model assumes that the ventral
stream is largely bilaterally organized and that the
dorsal stream is strongly left-hemisphere dominant.
The bi-hemispheric representation of the ventral
stream has to be understood as an asymmetrical
distribution of computational processes. The evidence
from brain lesion studies, supports the notion that
lesions of the regions involved in the right ventral
stream do not disturb language. These regions when
undamaged can in turn compensate for aphasia
resulting from lesions on the left side. This was
already suggested in the models of Wernicke and
we were also able to demonstrate this effect with
dichotic listening in the recovering from aphasia
(Castro-Caldas and Silveira Botelho,1980).
It is necessary, however, to consider that the
findings that support this model are the result
of the adaptation of a certain kind of brain to a
certain kind of stimulation. There are different
brains and different stimuli (for a general review
see, for instance, Coppens et al Eds.,1998). Deaf
subjects are an example of an exception. We know
very little about the importance of the absence of
hearing in fetal brain development, we know for
sure that the brain is not stimulated by sound and
we know that communication stems on visuo-motor
system different from the audio-motor system of
hearing subjects. Therefore all the considerations
that fed the model mentioned above need to be
reviewed a least on what concerns the ventral
stream. However, we know the brain cortex processes operations that are modality independent.
This means that the temporal cortex processes
auditory information not because the information
is based on sound but because the structure of
the information that is carried is best suited to be
processed there. This is also true for vision: when
born blind subjects read by Braille they activate
the visual cortex (Sadato et al. 1998).
It is still important to consider that, from the
evolutionary point of view, language is probably the
result of the evolution of a macaque mirror neuron
system for action perception and production (Arbib,
* [email protected]
Cadernos de Saúde  Vol. 2  Número especial de Línguas Gestuais – pp. 7-12
8
2005). Indeed, since the seminal report of Rizzolatti
in monkeys (Rizzolatti et al, 1996) we accumulated
evidence of the importance of this system for human
cognition and behavior.
Through this system both hearing and deaf subjects
acquire the competence of communicating which
is obviously paced by the cognitive competence
of pragmatics. The difference is that deaf subjects
use vision to receive information and the hands to
produce communicative motor actions and hearing
subjects can use two systems. For hearing subjects
there is the natural preference of receiving acoustic
information and producing communicative motor
actions with the vocal articulatory system.
The first question that we can raise, therefore,
concerns plasticity. We know that the comparison
between schooled and unschooled subjects makes
a difference in the development of the brain both
functionally and anatomically. The comparison made
between these two groups in adult age showed
functional differences while subjects repeted words
and pseudo-words (Castro-Caldas et al, 1998) and
anatomical differences in connectivity through the
corpus callosum (Castro-Caldas et al, 1999).
With this propose Allen et al (2008) investigated
whether auditory deprivation and/or sign language
exposure during development alters the macroscopic
neuroanatomy of the human insula. Volumetric
analyses were based on MRI data from 25 congenitally deaf subjects who were native users of
American Sign Language (ASL), 25 hearing subjects
with no knowledge of ASL, and 16 hearing subjects
who grew up in deaf families and were native
ASL signers. Significant variation in insula volume
was associated with both hearing status and sign
language experience. Compared with both hearing groups, deaf subjects exhibited a significant
increase in the amount of gray matter in the left
posterior insular lobule, which they hypothesize
may be related to the dependence on lip-reading
and articulatory-based (rather than auditory-based)
representations of speech for deaf individuals. Both
deaf and hearing signers exhibited an increased
volume of white matter in the right insula compared with hearing nonsigners. They hypothesize
that the distinct morphology of the right insula for
ASL signers may arise from enhanced connectivity
resulting from an increased reliance on cross-modal
sensory integration in sign language compared with
spoken language.
This study illustrates well the plasticy of the brain
related to the type of information processing. Other
differences are probably present that need to be
acknowledged along with the better understanding
of the computational processes involved both in
oral and in sign language.
We have to be careful, however, when studying
the correlation between function and brain organization in sign language. There are different ways
of processing sign language: the subject may be
born deaf from deaf parents and be raised very
early through sign language, or he can be deaf and
born from hearing parents that learn sign language
to communicate with him, or he can hear being
born from deaf parents and learn very early sign
language, or sign language can be learned by hearing
subjects in adult life. All this and other possibilities
may have an expression in the areas of the brain
that are recruited to fulfill the function. Therefore
the subjects to be included in the different studies
need to be well selected and the results obtained
should not be generalized.
The last general question that we think it is
necessary to have in mind before trying to find
the biological basis of sign language concerns the
proper nature of the communication system or, in
other words, is sign language a language that parallels
oral language in all its features or is it a different
competence. Not being a linguist it is hard to make
a deep analysis of this topic therefore I prefer to
quote Klima and Bellugi (1979) that in there seminal
work about American Sign Language wrote: “We do
not mean to argue that spoken language and sign
language are essentially the same. Certainly we would
be the last to argue that speech does not constitute
part of the biological foundations of language. But
if speech is specially selected, if sound constitutes
such a natural signal for language, then it is all the
more striking how the human mind, when deprived
of the faculty that makes sound accessible, seizes
on, perfects, and systematizes an alternate form to
enable the deeper linguistic faculties to give explicit
expression to ideas”(p.315). Furthermore we have
still to consider that as we have differences in the
structure of oral languages around the world, which
matters for mind-brain correlations, it is possible
that such differences are also present among sign
languages.
Lesion Studies
The classical method to interpret brain function
was based on the observation of patients that for
some reasons acquired a brain lesion. There are
several theoretical arguments against this method.
The main argument is also a classical one and
was raised by Jackson in the XIX century. Jackson
claimed that “to locate the damage which destroys
speech and to locate speech are two different things”
(apud Head, 1926). Other arguments are related to
the proper nature of the cerebral lesions. Vascular
lesions are the more frequent case and were the
basis for most of the knowledge that was acquired
for the past century. However, ischemic lesions are
not random because they occur in vascular territories
that are similar from subject to subject. When we
discuss vascular aphasia we are talking about arterial
syndromes. Traumatic lesions are usually multiple
and therefore they are difficult to localize. Tumors
are also responsible for focal syndromes, however
they are progressive lesions and there are always
progressive recovery mechanisms accompanying the
growing process of the tumor.
Even though, we learned very early with Broca
that language was dependent on left hemisphere
mechanisms (Broca, 1865) and with Benson that
fluent aphasias were the result of post-central lesions
and non-fluent aphasias were the result of precentral lesions (Benson, 1967). Both these aspects
are important land markers in the history brain
function. Therefore, studying patients with brain
lesions is still a fundamental source of evidence that
contributes enormously to our knowledge.
The work of Poizner, Klima and Bellugi (1987) is
unavoidable to discuss the effects of brain lesions
on sign language performance of deaf subjects.
They report their findings in patients with both right
and left hemisphere lesions. These and subsequent
findings aloud the conclusion that sign language was
also left-lateralized and that the classical localizations of lesions resulted in comparable dysfunction
between oral and sign speakers. It was suggested
that the visual component of sign language could
be an important aspect to consider. As a matter of
fact signs related to space seem to be disturbed by
right hemisphere lesions.
Reviewing the literature on aphasia in deaf
subjects, Corina (1989) considered that there is a
clear indication of left hemisphere dominance for
language. There were also findings supporting the
relationship of anterior and posterior lesions of
the left hemisphere with deficits in producing and
deficits in comprehension. “However, this author
stated (p.37), whether the exact neural substrates
underlying the symbolic systems supporting language
comprehension and production are shared by spoken
9
and signed language remains to be determined”.
He quotes several single cases in which the comprehension deficit was not related to the classic
lesion in Wernicke’s area, and cases in which the
production deficit was due to lesions away from
Broca’s area.
In this paper the author also reviews some
aspects of neurolinguistics and the role of the right
hemisphere. In the first case the disturbances in
sign productions seem to parallel those found in
oral aphasia, like the presence of paraphasias of
different types or agramatism. Even jargonaphasia
was reported in a case from 1943 (Leischner, 1943).
The question of the right hemisphere seems to
be more complex. Corina (1989) considered that “it
seems reasonable to entertain the possibility that the
right hemisphere damage does not disrupt linguistic
function per se, but rather impairs the execution
and processing of linguistic information in sign
language, in which spatial information plays a
particularly salient role. However, the issues become
more complicated when we consider the syntactic
aspects of ASL.” There is evidence from some of
the cases with right hemisphere lesions that the
spatialized syntax is disturbed and Poizner et al
(1987) considered that the perceptual processing
involved in the comprehension of spatialized syntax
involves both left and right hemispheres.
Studies on the development of the brain of hearing
children revealed that there is a correlation of the
growing pattern of the cortex of the inferior frontal
grey matter of the left hemisphere and the development of phonology (Lu et al, 2007). On the other
hand, we know that the absence of stimulation is a
reason for the non development of language (see,
for instance, Curtis,1977). Therefore there is a crucial
period in which the stimulation induces learning
and brain development. The study by Newman et
al (2002) calls the attention for a critical period for
right hemisphere recruitment in ASL processing. The
authors found that the right angular gyrus was active
during ASL processing only in those subjects that
were native hearing bilinguals (ASL-English) and not
in late learners of ASL. This finding is crucial for
teaching processes: if the brain is stimulated during
the critical period it develops the best structures to
deal with the information and therefore we may
expect that the quality of the processing is better.
It is well known that learning a second language
late in life is possible but is very rare that this is
done without a foreign accent.
10
The link between findings on activation studies
with lesion studies is not always easy to establish.
Even concerning the involvement of each hemisphere
in sign language comprehension Hickok et al (2002)
studied a group of nineteen lifelong signers with
unilateral brain lesions. The subjects were tested on
comprehension with different degrees of complexity.
Results showed that subjects with left temporal
lobe lesions were much more impaired than the
remaining subjects and subjects with right hemisphere
lesions were as impaired as hearing subjects with
right hemisphere lesions. The authors concluded
that deaf and hearing individuals have a similar
degree of lateralization of language comprehension
processes and that language comprehension in sign
language depends primarily on the integrity of the
left temporal lobe.
Activation Studies
Contrary to lesion studies in which Nature decides
the localization of the lesion and the observer tries
to reveal the processing rules by finding what is
disturbed and what is preserved in the performance
of patients, in activation studies the observer has to
design a task directed to a computational operation
in order to reveal which region of the brain becomes
active.
For the past decade a great number of experimental work has been published illustrating several steps
of the complexity of the mechanisms responsible for
comprehension and production of sign language. It
is hard for the moment to make a comprehensive
review of this literature. Therefore, I will review
some of the work that can be considered more
seminal.
May be we could start a so-called Cartesian question. Movement, which is a quality of the body and
cognition, which is a quality of the mind. Willems
and Hagoort (2007) reviewed recently this question focusing mainly on co-speech gestures. As the
authors mentioned, co-speech gestures embody a
form of manual action that is coupled to the language
system. Both gestures and spoken language occur
simultaneously with the intention of communicating.
Therefore they belong to one integrated system of
communication (see McNeill, 1992).
As it was mentioned above both intentional gestures and speech sounds activate the mirror neuron
system. When subjects listen to meaningful syllables
there are significantly more activity in bilateral precentral gyrus and central sulci than when subjects
listen to meaningless sounds (Wilson et al 2004).
If we go back to work of Petitto and Marentette
(1991) that suggested the parallel development of
oral and hand babbling as precursors of oral and
sign language we can admit that for the general
intention of communicating the brain of deaf signers
and oral speakers recruits the same regions.
More recently, Corina et al (2007) studied in more
detail the activity of this mirror neuron system while
both deaf signers and oral speakers were viewing
different actions. Subjects were scanned in three
different conditions: intransitive self-oriented actions,
transitive object-oriented actions and symbolic actionsigns used in ASL. Previous research suggested the
involvement of bi-lateral prefrontal cortex (PFC),
parietal and superior temporal sulci (STS) activation
related to this mirror neuron system (Grezes and
Decety, 2001) in hearing subjects. Some authors
suggested that it was possible to postulate a model
of sensory motor control involving the STS, PFC
and F5 in the perception of action and a reverse
model used to generate predictions of movement
outcome during imitated actions (Miall, 2003; Carr et
al, 2003; Iacoboni, 2005). The consistency of these
hypothesis in still to be definitively understood and
it is well accepted that the nature of the gesture
is an important factor to make differences in the
activation patterns.
The results of the study by Corina et al (2007)
demonstrated that the different classes of actions
engaged the a frontal/parietal/STS human action
recognizing system in an highly similar fashion. This
neural consistency across motion classes was true
primarily for hearing subjects. Deaf signers engaged
left-hemisphere perisylvian language areas during
the perception of signed language signs. They also
did not engage the expected fronto/parietal/STS
circuitry during passive viewing of non-linguistic
actions but instead they activated middle-occipital
temporal-ventral regions which are known to participate in the detection of human bodies, faces and
movements. The authors suggested that deaf subjects
may engage specialized neural systems that allow for
rapid online differentiation of meaningful linguistic
actions from non-linguistic human movements.
The study from Campbell and Capek (2008) brings
new insight to the knowledge of the mechanisms
that are involved in visual processing of information
related to communication. The authors raised two
questions: do deaf and hearing people differ in
the regions activated by (silent) speechreading?,
and how does the presence of mouth actions in
the sign affect the cortical activation pattern? A
group of deaf subjects with good proficiency in
lip-reading was compared to a group of hearing
subjects while viewing blocks of silent spoken
words in the fMRI. For the first question the authors
found that the left superior temporal cortex, including auditory regions, was strongly activated in the
brains of deaf compared with hearing participants
when processing silent spoken (speechread) word
lists. In the second place, they found that within
the signed language, cortical activation patterns
reflected the presence and type of mouth action
that accompanied the manual sign. Signed items
that incorporated oral as well as manual actions
were distinguished from signs using only manual
actions. Signs that used speechlike oral actions
could be differentiated from those that did not.
Thus, whether in speechreading or in sign language
processing, speechlike mouth actions differentially
activated regions of the superior temporal lobe that
are accounted auditory association cortex in hearing
people. This is a good example of the plasticity of
the brain as was mentioned above.
As it was mentioned above the neurobiology
of sign language (and even the neurobiology of
spoken language) is far from being a finished
discussion. The recent paper by MacSweeney et al
(2008) calls the attention to some of the outstanding
questions that need clarification and that are a
good way of concluding this brief review and are
as follows: “1. What is the influence of age of first
language acquisition on language processing and
its underlying neural systems? 2. How do memory
and language systems interact and what is the
impact of language modality? 3.To date, iconicity
does not seem to influence SL processing. However,
might iconicity have a role in semantic organization and imagery? 4. What characteristics of SL
and SpL influence laterality of processing? 5. Can
studies of SL give us further insights into the extent
to which the mirror neuron system is involved in
language processing? Are the fronto parietal parts
of this system differentially involved in SL, SpL and
gesture processing? 6. To what extent are regions
associated with language processing driven by the
intentional stance of the observer and the potentially
communicative nature of the actions? 7. Speakers
can hear themselves speak. Signers do not see their
own signing as others see them. What are the roles of
visual and proprioceptive feedback and the proposed
mirror system in language monitoring? 8. Does the
mouth have a gestural role in SL similar to that of
11
the hands in SpL? Are these “gestures” processed
similarly in the brain? 9. How do links between
hand and vocal gestures inform our understanding
of the evolution of language and its neurobiological
bases?” (p.438).
References
1.
Allen, J.S, Emmorey, K., Bruss, J. and Damasio, H.(2005). Morphology of
the Insula in Relation to Hearing Status and Sign Language Experience.
The Journal of Neuroscience, 28 :11900-11905.
2.
Arbib, M.A. (2005). From monkey-like action recognition to human language: An evolutionary framework for neuroliguistics (with commentaries
and author’s response). Behavioral and Brain Sciences, 28: 105-167.
3.
Benson, D. F. (1067). Fluency in aphasia: correlation with radioactive scan
localization. Cortex, 3: 373-394.
4.
Broca, P. (1865). Du siege de la faculté du langage articulé. Bulletin de la
Société d’Anthropologie, 6: 373-393.
5.
Campbell, R. and Capek, C. (2008). Seeing speech and seeing sign: Insights
from a fMRI study. International Journal of Audiology, 47.1,S3-S9.
6.
Carr, L., Iacoboni, M., Dubeau, M.C., Mazziotta, J.C. and Lenzi, G.L. (2003).
Neural mechanisms of empathy in humans : a relay from neural systems
for imitation to limbic areas. Proc. Natl. Acad. Sci. U.S.A.,100: 5497-5502.
7.
Castro-Caldas, A. and Silveira Botelho, M. A (1980).Dichotic listening in
the recovery of aphasia after stroke. Brain and Language, 10: 145-151.
8.
Castro-Caldas, A., Petersson, K.M., Reis, A., Stone-Elander, S. & Ingvar, M.
(1998). The illiterate brain: learning to read and write during childhood
influences the functional organization of the adult brain. Brain, 121:
1053-1063.
9.
Castro-Caldas, A., Miranda, P., Carmo, I., Reis, A., Leote, F., Ribeiro, C. &
Ducla-Soares, E. (1999). Influence of learning to read and write on the
morphology of the Corpus Callosum. European Journal of Neurology, 6:
23-28.
10. Coppens, P., Lebrun, Y. and Basso, A. (Eds.) (1998) Aphasia in Atypical
Populations. Lawrence Erlbaum Associates. New Jersey.
11. Corina, D. P. (1998). Studies of Neural Processing in Deaf Signers: Toward
a Neurocognitive Model of Language Processing in the Deaf. Journal of
Deaf Studies and Deaf Education, 3: 35-48
12. Corina, D., Chiu, Y.-S., Knapp, H., Greenwald, R., San Jose-Robertson, L.
and Braun, A.(2007). Neural correlates of human action observation in
hearing and deaf subjects. Brain Research, 1152: 111-129.
13. Grezes, J. and Decety, J. (2001). Functional anatomy of execution, mental
simulation, observation, and verb generation of actions: a mete analysis.
Hum. Brain Mapp., 12: 1-19.
14. Head, H. (1926). Aphasia and kindred disorders of speech. Cambridge
University Press. UK.
15. Hickok, G., Love-Geffen, T. and Klima, E. S. (2002). Role of the left hemisphere in sign language comprehension. Brain and Language, 82:167-178.
16. Hickok, G. and Poeppel, D.(2007). The cortical organization of speech
processing. Nature Reviews Neuroscience, 8: 393-402.
17. Iacoboni, M. (2005). Neural mechanisms of imitation. Curr. Opin. Neurobiol.,
5: 632-637.
18. Klima, E. and Bellugi, U. (1979). The signs of language. Harvard University
Press, Cambridge Mass.
19. Leischner, A. (1943). Die “Aphasie” der Taubstummen. Archiv fur Psychiatrie
und Nervenkrankheiten, 115: 469-548.
20. Lu, L.H., Leonard, C. M., Thompson, P. M., Kan, E., Jolley, J., Welcome, S. E.,
Toga, A. W. and Sowell, E.R. (2007). Normal Developmental Changes in
Inferior Frontal Gray Matter Are Associated with Improvement in Phonological Processing: A Longitudinal MRI Analysis. Cerebral Cortex,17:1092-1099.
12
21. MacSweeney, M., Capek, C.M., Campbell, R. and Woll, B. (2008). The signing
brain: the neurobiology of sign language. Trends in Cognitive sciences,
12:432-440.
22. McNeil, D. (1992). Hand and Mind: What gestures reveal about thought.
University of Chicago Press. IL.
23. Miall, R. C. (2003).Connecting mirror neurons and forward models. NeuroReport, 14: 2135-2137
24. Newman, A.J., Bavelier, D., Corina, D., Jezzard, P. and Neville, H.J. (2002). A
critical period for right hemisphere recruitment I American Sign Language
processing. Nature Neuroscience, 5: 76-80.
25. Poizner,H., Klima, E. and Bellugi, U. (1987). What the Hands Reveal About
the Brain. The MIT press, Cambridge, Mass.
26. Petitto, L. and Marentette, P. (1991). Babbling in the manual mode: evidence for the ontogeny of language. Science, 251: 1493-1496.
27. Rizzolatti, G., Fadiga, L., Gallese, V. and Fogassi, L. (1996). Premotor cortex
and the recognition of motor actions. Cognitive Brain Research, 3: 131-141.
28. Sadato, N., Pascual-Leone, A., Grafman, J., Deiber, M.P., Ibanez, V. and
Hallett, M.(1998). Neural networks for Braille reading by the blind. Brain,
121: 1213-1229.
29. Willems, R. M. and Hagoort, P. (2007). Neural evidence for the interplay
between language, gesture, and action: A review. Brain and Language,
101: 278-289.
30. Wilson, S. M. Saygin, A. P., Sereno, M. I. and Iacoboni, M. (2004) Listen
to speech activates motor areas involved in speech production. Nature
Neuroscience, 7: 701-702.
Phonological category resolution:
A study of handshapes in younger and older sign languages*
Categorias fonológicas:
um estudo sobre orientação manual em línguas gestuais antigas e novas
Assaf Israel1** and Wendy Sandler1***
1
University of Haifa / Sign Language Research Lab
Resumo
Abstract
A existência de um sistema fonológico – um sistema em
que se articulam unidades com um significado mínimo para a
criação de palavras significantes – é, muitas vezes, considerada
como um pré-requisito para a linguagem. A descoberta de que
as línguas gestuais contêm um nível de estrutura significante
convenceu, definitivamente, os linguistas de que tratava de
verdadeiras línguas.
Contudo, a questão da emergência do sistema fonológico
não foi ainda tratada, tendo em conta que, por um lado, as
línguas orais são línguas antigas ou descendentes de línguas
antigas e, por outro lado, a maior parte das línguas gestuais
que foram estudadas até aos dias de hoje, já têm algum tempo
de existência.
O presente estudo constitui-se como um passo que documenta a formação de categorias fonológicas numa nova língua
gestual: a Língua Gestual Al-Sayid Beduína (ABSL). Esta nova
língua emergiu recentemente numa comunidade isolada com
uma grande incidência de Surdez.
Este trabalho foi motivado pela observação de que esta
nova língua parece exibir uma variação na formação de gestos
entre os gestuantes (Aronoff et al. 2008). Para dar conta deste
fenómeno, medimos a variação de 10 gestuantes relativamente
à produção de uma categoria fonológica – configuração – em 15
gestos de ABSL, comparando os resultados com a produção
da configuração em duas outras línguas mais estabelecidas – a
ISL (Língua Gestual Israelita) e a ASl (Língua Gestual Americana).
A nossa metodologia mede o grau de consenso entre os
gestuantes, relativamente a cada característica fonética da
configuração e ao número de variantes exibido, revelando um
padrão consistente nas três línguas em estudo.
A maior quantidade de variação foi encontrada em ABSL,
seguida da ISL. A ASL foi a língua que apresentou menos variação na produção da categoria configuração. Estes resultados
sugerem que as categorias fonológicas ainda se encontram em
processo de formação numa nova língua e acreditamos que
a combinação de factores históricos e sociais podem explicar
a gradação exibida ABSL > ISL > ASL. Os nossos achados
The existence of a phonological system – a system of
meaningless building blocks that make up meaningful words – is
often considered a prerequisite for language, and the discovery
that sign languages used by deaf people have a meaningless
level of structure convinced linguists that they are real languages.
But the question of how a phonological system arises has
not previously been addressed, since all spoken languages are
old or descended from old languages, and most sign languages
that have been studied have also been around for some time.
The present study is a step toward documentation of the
formation of phonological categories in a new sign language,
Al-Sayyid Bedouin Sign Language (ABSL), which arose recently
in an insular community with a high incidence of deafness.
The work is motivated by the observation that this new
language appears to exhibit a good deal of cross-signer variation
in the formation of signs (Aronoff et al 2008).
To put this observation to the test, we measure the amount of
variation across 10 signers in the production of one phonological
category – handshape – in 15 signs of ABSL, and compare
it with handshape production in two other, more established
sign languages: Israeli Sign Language (ISL) and American Sign
Language (ASL).
Our methodology measures the degree of cross-signer
consensus with respect to each meaningless phonetic feature
of handshape as well as the number of variants (indicating the
range of variation), and reveals a consistent pattern across the
three languages:
The largest amount of variation is found in ABSL; ISL is next;
and ASL shows the least amount of cross-signer variation in
production of the handshape category.
These results suggest that phonological categories are
still in the process of being formed in the new language, and
we appeal to a combination of historical and social factors to
explain this ABSL > ISL > ASL cline.
The findings and analysis offer a glimpse into the development
of phonological categories in a new language.
* This work was supported by the United States-Israel Binational Science Foundation and the National Institute on Deafness and
other Communication Disorders of the National Institutes of Health.
* This paper also appears in Rachel Channon & Harry van der Hulst (eds.). Formational Units in Sign Language. Ishara Press.
** [email protected]
*** [email protected]
14
e a sua análise oferecem um primeiro olhar relativamente ao
desenvolvimento das categorias fonológicas numa nova língua.
Keywords: Sign language phonology, phonetic variation,
Al-Sayyid Bedouin Sign Language, language emergence 
Palavras Chave: Fonologia das Línguas Gestuais, variação
fonética, Lingua Gestual Beduína AL – Sayyid, emergência da
linguagem 
Linguists began to take sign languages seriously
as an object of study after William Stokoe demonstrated that the words of these visual languages are
constructed from a discrete and finite list of meaningless units – that they have phonology (Stokoe,
1960). This discovery dispelled the myth that sign
languages were comprised of iconic gestures, holistic
in form. But do those meaningless units that Stokoe
identified exist in a new sign language? Or does it
take time for a phonological system to self-organize?
Investigating a new sign language gives us insight
into this question, and to the broader question of
whether it is necessary to have phonology in order
to have language.
The myth that established sign languages like
American Sign Language (ASL) were holistic gesture
systems was based on iconic properties that are
readily observable in many signs: A sign for book
looks like opening a book and a sign for eat looks
like putting food in the mouth, to take examples
shown in Figure 1 from Israeli Sign Language (ISL).
Figure 1 – BOOK, EAT, and LEARN/STUDY in ISL.
But Stokoe showed that the handshapes, locations,
and movements of signs perform the same role in
the lexicons of sign languages as do the meaningless
sounds that make up spoken words, coming together
in different combinations to create new lexical items.
So, ISL LEARN is distinguished from EAT by being
articulated at a different location, and ISL ASK and
SAY in Figure 2 are a minimal pair, distinguished
by aspects of handshape.
Stokoe posited handshape, location, and movement as the three major categories of meaningless
formational elements in ASL, each with a finite list of
contrastive formational elements, to which we return
in Section 2. Since then, a good deal of evidence has
Figure 2 – ISL ASK and SAY, minimally distinguished by handshape
accrued for the existence of a phonological level of
structure in sign languages, one which consists of
categories and features (Liddell & Johnson, 1989), as
well as constraints on their form and combination
(e.g., Mandel, 191,; Battison, 198,; Sandler, 199,;
Corina, 193,; Brentari, 198,; Sandler, 1999). Further,
researchers found evidence for hierarchical organization of feature classes based on their behavior in
assimilation and other phenomena (Sandler, 197,;
1989; Corind & Sagey, 1989). These constraints and
processes hold on elements of form, irrespective
of meaning, showing that sign languages have a
meaningful and a meaningless level of structure, a
characteristic that Hockett (1960) called ‘duality of
patterning’ and proposed as a basic design feature
of human language.
Although it is assumed that all sign languages have
iconic roots – and it would be an inefficient visual
language indeed that did not take advantage of this
possibility – it has been shown that diachronically
signs become less and less iconic (Frishberg, 1995).
Over time, they become more restricted and symmetrical, and signs that once involved other parts
of the body came to be represented as symbolic
images conveyed only by the hands. The changes
are in the direction of the self organization of a
formal system of meaningless units.
But how does this process take place? More specifically, how does a language develop phonological
categories? The answer to this question cannot be
discovered empirically in spoken languages, as they
are all very old. Even pidgin speakers have full
command of the phonology of their millennia-old
native languages. But sign languages arise anew
whenever the right conditions are met – whenever a
group of deaf people have an opportunity to gather
Phonological category resolution
and communicate regularly. And as new languages,
sign languages have much to teach us about the
emergence of linguistic form.
Here we examine in detail the formational characteristics of signs in a new language – Al-Sayyid
Bedouin Sign Language (ABSL) – and compare them
to those of two other sign languages with different
social histories, American Sign Langauge and Israeli
Sign Language. Initial observation of vocabulary
items signed by different people across the Al-Sayyid
village revealed unexpected variation – both in the
choice of lexical items and in the form of the same
lexical item. Following up on this observation, we
conducted a detailed analysis of the form of sign
productions in isolation.
We focus on handshape in the present study,
and describe our investigation, which confirmed
our impression of considerable variation along most
of the parameters involved in this category. By
comparing sign productions with those of two other
sign languages, we see a cline, with ABSL exhibiting
the most variation in the formation of handshapes,
ISL next, and ASL showing the least variation across
signers. Taken together with other evidence, we
hypothesize that ABSL signers are aiming for a
holistic iconic image, and that discrete phonological
categories are not yet robust in the language.
We begin with a description of ABSL in Section
1, and illustrate with some of the variation in sign
production that was discovered in the broader ABSL
research project (Aronoff et al., 2008; Sandler et al.,
2009). We then turn to the study of handshape,
coding and analyzing handshapes in 15 signs for
ten signers in each of the three languages. Section
2 describes the handshape features of interest and
the methodology is the topic of §3. Results and
discussion follow, in §4. Alongside the variation in
sign production, the ABSL team has also observed
early indications of formal organization, and we
describe some of these in this section. In Section
5, we consider some explanations for differences
across languages, including language age, community size and other social characteristics. While our
results suggest that ABSL has not yet formed discrete
phonological categories, we see some evidence
pointing in that direction, which we exemplify in
Section 6. Section 7 is a summary and conclusion.
1. Al-Sayyid Bedouin Sign Language
The only languages that arise de novo with no
model are sign languages, and we have much to learn
15
by observing their early evolution. A sign language
emerges whenever a community of deaf individuals
is formed, and there are two different routes through
which this happens (Meir et al., 2009). A common
route is through establishment of schools for deaf
children, where local sign languages (and sometimes
foreign sign languages like French Sign Language in
the case of ASL) together with home sign systems
mingle to give rise to national sign languages. Most
of the sign languages that have been well studied
are deaf community languages of this kind, formed
within the past 75 to 300 years. Another setting in
which sign languages develop is that of relatively
isolated communities with higher than average rates
of deafness, where village sign languages are born.
Meir et al. (2009) describe six village sign languages
in different parts of the world in their survey of new
sign languages, but there are many more.
The best known deaf community sign language
that is new is Nicaraguan Sign Language (NSL),
forged from home sign systems when deaf children
were first brought together in a school in Managua
in 1977. Research on this language has shown that
systematic language structure arises when children
were brought to the school at a young age, with
older children using a more idiosyncratic system
as a language model. Spatial modulation – the use
of space to indicate the different grammatical roles
in a sentence – is one way in which NSL gradually
became more systematic at the morpho-syntactic level
(Senghas, Coppola & Newport, 1997; Senghas, 2003).
The present study focuses on a young village
sign language, ABSL. The language took root in the
Al-Sayyid Bedouin village in the Negev Desert of
present day Israel, when four deaf children were born
in a single household about 75 years ago. Due to
its insular social structure, consanguineous marriage
patterns and high birth rate, genetic deafness spread
in the population (Scott et al., 1995), and today,
there are about 120-150 deaf people in the village.
An indigenous sign language arose among the deaf
people and is used by many of the hearing villagers
as well (Kisch, 2000).
ABSL functions as a full fledged language, used
for a range of social interactions, for instructions and
plans, and to discuss such topics as personal histories,
folk remedies, national insurance, childcare, or how
to cajole a husband. The sentences of the second
generation of ABSL signers are verb-final, with SOV
word order in sentences with all three constituents,
and noun-modifier order in noun phrases (Sandler
et al., 2005).
16
While typical sign language morphological structures such as verb agreement and classifier predicate
constructions have not been found to exist in this
young language, a kind of size and shape classifier
affixation is common across the village (Meir et al.,
to appear; Sandler et al., 2009).
The ABSL research team began to learn about
the lexicon of this language as part of a dictionary
project in which several hundred lexical items were
recorded. This project had two surprises in store
for the team. One was the degree of variation in
lexical items themselves. Even signs for everyday
items sometimes had several variants. There is, of
course, a vocabulary of conventionalized signs, but
this conventionalization seemed to the researchers
to hold at the level of the overall image depicted by
the sign. Aronoff et al. (2008) observed that across
tokens produced by different signers there is variation
in sublexical components, which, according to the
authors’ impression, is greater than what they would
expect in more established sign languages, such as
ISL and ASL. Moreover, tokens seem to vary across
features that are potentially contrastive in established
sign language. One example is variation in place
of articulation found in different tokens for ABSL
DOG. The example is repeated in Figure 3.
of subgroups within a language community. Rather,
it seems to us that this variation is an indication that
the ABSL lexicon has not yet developed discrete,
meaningless formational categories. No minimal pairs
have surfaced in all the recorded forms. In other
words, we hypothesize that it takes time for users
of a new language to converge on a relatively fixed
set of primitives for forming lexical items.2
In fact, closer examination of productions of the
sign DOG shows variation along all parameters,
as reflected in Table 1, from Sandler et al. (2009).
Table 1 – variation across sublexical parameters in different
productions of the ABSL sign DOG.
Figure 3 – Variation in location across ABSL tokens for DOG.
(a) DOG
(b) DOG
The variant in (a) is articulated in neutral space and
the variant in (b) is signed in front of the mouth.1
Variation in ABSL such as illustrated above may
be compared to variation in the pronunciation of the
English word route, [rut] and [raut]. In the English
example, we tend to associate the variation with
different varieties or dialects. However, in the case of
ABSL, the different signers whom we have recorded
are members of the same extended family within a
small, closely-knit community, and we suspect that
variation is not ‘sociolinguistic’ in the normal sense
1
The figures are taken from Aronoff et al. (2008).
In order to test the hypothesis that distinct formational categories are not yet defined in ABSL, it
is necessary to record and analyze the amount of
variation at different points along the development
of a language. We expect that the ongoing research
on ABSL will provide us with insights into this issue
as the language develops further. At present, we
choose to use other, more established languages as
points of reference against which variation in ABSL
may be compared.
2
Models of linguistic communication proposed in a number of
computational studies produce gradual convergence across
different “language users” (see, for example, Barr, 2004;
Hutchins & Hazlehurst, 1995)
17
The current study, taken from a larger project
(Israel, 2009), focuses on one sublexical component
– hand configuration. In the next section, we discuss
briefly the internal structure of this component as a
way of introducing the phonetic features that we will
use for the coding of signs. This will be followed
by a discussion on the measures of variation and
the way to compare them across languages.
2. Sublexical structure in sign language:
hand Configuration
As noted in the introduction, the most influential
study in the field of sign linguistics was William
Stokoe’s analysis of the internal structure of signs in
ASL. Stokoe (1960) was the first to show that signs
could be broken into sublexical components, much
as spoken words are analyzed as combinations of
different sounds. In his analysis, Stokoe referred
to three different parameters – hand configuration,
location and movement – whose different specifications were proposed to be sign language analogues
of phonemes. In other words, Stokoe showed that
contrasts between different signs were made by
substituting one sublexical component for another,
similar to the way spoken words are distinguished
by different consonants and vowels. Figures 1 and
2 above showed minimal pairs for location (the
mouth for EAT and the forehead for LEARN) and
for hand configuration ( in ASK and
in SAY)
in a different sign language, ISL.
Stokoe’s unraveling of the systematic patterning
of sublexical components in ASL led to widespread
recognition of sign languages as bona fide human
languages. That recognition motivated extensive
research of sign language structure at all levels of
organization. According to more recent accounts,
noted above, sublexical components are organized
in more complex structures, and each component
has its own internal structure. In this study, we are
concerned with variation in one complex component
– hand configuration.3 We begin by discussing the
internal structure of hand configuration. The purpose
of the discussion is to arrive at a set of features
that may be used for a transparent coding of hand
configuration which can capture phonetic variation.
A considerable portion of the sign language phonology literature has been dedicated to the lexical
3
The full study compares variation across the three languages
in all three major categories, location and movement in
addition to handshape (Israel, 2009).
representation of hand configuration. Although the
specific representation of this component is different
across models, there is consensus among phonologists about certain generalizations that should
be captured. A fundamental distinction is made
between finger selection and finger position: while
the position of the fingers may change within a sign,
the selection of fingers remains fixed (Mandel, 1981),
unless the sign is multimorphemic, in which case
finger specification may change across morphemes
(Sandler, 1989). This distinction is maintained in
several models (Sandler, 1989; van der Hulst, 1993;
van der Kooij, 2002; Brentari, 1998).
The first complex model of hand configuration
was proposed by Sandler (1987; 1989). In that
model, finger selection was represented by a set
of five features – one for each finger. That is, if
a certain handshape selects the index and middle
fingers, then the underlying specification of finger
selection for this handshape is [+thumb/+index].
This transparent form of representation comes at
the expense of economy and explanatory power
(cf. later proposals by Sandler 1995; 1996 and van
der Kooij, 2002), but it serves well the purpose of
representing phonetic variation, the current focus.
In the coding of finger selection for the analysis of
variation described below, we shall use the unary
features [index], [middle], [ring], [pinky] and [thumb].
As for the position of the fingers, we distinguish
between the four different degrees of flexion illustrated in Figure 4. The hand’s anatomy allows for
flexion at the Metacarpophalangeal joints (or, ‘base
joints’) only, at the proximal and distal Interphalangeal joints (also referred to as the ‘non-base joints’)
only, or at both base and non-base joints at the
same time (Ann, 1996). The four positions in Figure
4 are examples for the four combinations of binary
specifications for each type of joints.
Figure 4 – Four flexion positions in handshapes with all the
fingers selected.
The [extended] position has no flexion of the fingers; in a [bent]
position, only the base joints are flexed; in a [clawed] position only
the non-base joints are flexed; and in the [curved] position, base
and non-base joints are flexed.
[extended]
[bent]
[clawed]
[curved]
18
A generalization which concerns the position of
the unselected fingers was stated as the Unselected
Fingers Constraint (Corina, 1993) in (1).
1. Unselected Fingers Redundancy Rule:
If the selected fingers are in a closed position,
then the unselected fingers are open, as in
and ; otherwise, the unselected fingers are
usually closed, as in
and .
This generalization means that an underlying
specification of the unselected fingers’ position
would be redundant. Still, there are cases in which
the position of the unselected fingers is not predictable. For example, handshapes with contact
between the thumb and the index finger as the
only selected finger sometimes vary across signers,
with the unselected fingers kept closed by some
).
and open by others (
The position of the thumb is yet another feature
which is considered redundant. In handshapes with
less than all fingers selected, the thumb is often
crossed over the closed unselected fingers (Figure
5a). If the position of the selected finger(s) changes
from ‘open’ to ‘closed’ during the sign, the thumb
will often be in an ‘opposed’ position (Figure 5b).
In addition, some suggestions have been made
regarding an ‘extended’ position of the thumb (Figure
5c). Battison et al. (1975) indentified six features
associated with variation in the position of the thumb
in ASL signs with the index finger or the index
and middle fingers selected. Importantly, though,
those phonological environments did not fully predict thumb position, as different signers produced
different variants for the same signs. Battison et al.’s
interpretation of that variation is that it reflects a
process of diachronic change. Rules for the phonetic
implementation of thumb position have also been
proposed by van der Kooij (2002). In her model,
the ‘extended’ position of the thumb is realized a)
when the thumb itself is selected, or b) when the
selected fingers are spread.
Figure 5 – Thumb position features.
a. [crossed]
b. [opposed]
c. [extended]
Another hand configuration feature is finger
spreading. According to van der Kooij, a ‘spread’
position of the fingers tends to occur in dynamic
signs with all fingers selected, while in static signs
fingers are mostly adducted (or, ‘non-spread’). Again,
this is a generalization; phonological context does
not fully predict the surface form for this feature.
For example, one of our ISL consultants did not
accept a variant of the sign FISH – a dynamic sign
with all fingers selected – in which fingers were in
a spread rather than adducted position.
The last component of hand configuration to be
presented is orientation. This component is used
contrastively in sign languages that have been studied. An example from ISL of a minimal pair with
different hand orientations is shown in Figure 6.
Figure 6 – A minimal pair in ISL with two different orientations
of the hand.
COMPARE
VACILLATE
There is no consensus on how orientation should
be represented in phonological models of sign language. The option which seems to account best for
variation in this component refers to orientation in
relative terms, for example by indicating the part
of the hand which faces or contacts the place of
articulation or that which faces the direction of
movement (e.g., Mandel, 1981; Crasborn, 2001;
Crasborn & van der Kooij, 1997). However, since in
this study we are interested in capturing variation
Figure 7 – degrees of freedom in the representation of orientation.
The figure shows the position of the hand in the ISL sign SUNDAY.
The hand is oriented so that the metacarpal bones point upward
(specified as [up]), and the palm faces the contralateral side (specified as [contralateral]), as indicated with solid arrows.
in form, representing orientation in absolute terms
(i.e., in relation to the three dimensional space or
to the signer’s body) is more suitable. Therefore, we
shall specify orientation with the following features:
[up], [down], [in] (faces signer’s body), [out] (faces
away from signer), and [contralateral]. These features
will used to specify two degrees of freedom: the
side faced by the palm side of the hand and the
direction in which the hand’s metacarpal bones (i.e.,
the bones connecting the wrist and MCP joints)
point, as illustrated in Figure 7.
Table 2 summarizes the subcategories of hand
configuration and their respective features.
Table 2 – Hand Configuration subcategories and features.
Handshape
Selected Fingers
[index], [middle], [ring], [pinky], [thumb],
[any combination of fingers]
Flexion
[extended], [bent], [curved], [clawed]
Aperture
[open], [closed]
Spreading
[spread], [non-spread]
Unselected Fingers
[open], [closed]
Thumb
[extended], [opposed], [adducted],
[crossed]
Orientation
[up], [down], [in], [out], [contralateral]
As this study addresses convergence on the production of basic phonological elements, we are
not concerned here with models of the internal
organization of these features, which are based
partly on the behavior of phonological elements in
forms and rules.
3.Methodology
In order to determine whether there is indeed
more blurring of possible category distinctions in
ABSL than in other sign languages, we investigated
sign productions in three sign languages: ABSL, ISL,
and ASL. These languages have different histories
and social structures, which bear on the issue under
investigation. Ten signers from each group, each
signing 15 signs, provide the data fo the study
3.1. Participants
ABSL
Ten ABSL signers participated in the study. The
subject selection process took into consideration
social structure and constraints within the Al-Sayyid
community. Since deafness is genetically determined
19
in the village, and there is a good deal of first cousin
marriage, it is common for deafness to be particularly
widespread among close relatives, and indeed, the
signers included in the study are members of an
extended family, six of them members of the same
immediate family. The reliance on sign language
as the means of communication within the family
ensured that the signers chosen are highly proficient
in ABSL. A social constraint is imposed on women,
who cannot be videotaped if the recording might be
watched by men from the village. For those Al-Sayyid
women who participated in the study, consented to
participate with our reassurance that the recordings
would be used for analysis in the lab only. The
signers included in the study, then, are people who
are comfortable working with investigators and
are all highly proficient signers. There was a wide
distribution of ages among the subjects: two second
generation signers were between 40 and 50 years
old at the time of videotaping; one signer was about
28 years old – a young second generation signer; of
the third generation participants, fous were between
20 and 30 years od;, anst thres were between 7 and
12 years old. [ASSAF – SAY HOW MANY IN EACH
GENERATION AND AGES]
While many hearing people know ABSL well and
use it daily within deaf families, all ABSL participants
in the study were deaf. The oldest two, Th. and A-B.,
were born in the second generation of deaf people
in the Al-Sayyid village. The rest of the participants
represent the third generation of deaf people, all of
whom are attending school in a nearby village, where
some ISL signs are used by the teachers. At school,
children from Al-Sayyid interact with deaf children
from other villages in the area. However, neither the
majority of second generation deaf people nor the
hearing signers in the Al-Sayyid village have direct
exposure to ISL or other signs from other areas, and
we infer from this that the communication among
family members takes place in ABSL. In the study,
each signer signed to another ABSL signer while
being videotaped.
In order to compare variations across the three
sign languages, group sizes were balanced, so that
each of the ISL and ASL groups also numbered 10
signers each.
ISL
All 10 ISL participants were deaf signers. The ISL
group was formed in such a way that it would be
as analogous as possible to the ABSL group, both
linguistically and socially. Thus, all ISL participants
20
were from the same small geographic area, the
city of Haifa. This was intended to result in data
that are maximally lexically unified.4 In addition, of
the 10 participants, four were members of a single
immediate family (cf. the six same-family members
in ABSL), two of them one generation older (the
two parents), aged 45 and 50. The ages of the other
participants were 40, 38, 37, 32, 21, 21, and 14. All
ISL participants have had formal education.
ASL
The group of ASL signers was less homogenous
than the other two language groups. All 10 participants were “recruited” while spending leisure time
on the University of California, San Diego campus.5
All, except for a single participant (who acquired
ASL from a deaf parent), were deaf, and the only
requirement for participation was a perceived high
level of ASL proficiency. No information regarding
participants’ (original) area of residence, educational
background, etc. was collected. The ages of the
two youngest participants are between 20 and 30
years (the exact ages were not recorded). The other
participants were 32, 33, 35, 41, 42, 43 and 54 years
old. All the participants started learning to sign by
the age of six, with three participants acquiring sign
language from deaf parents.
3.2. Stimuli and procedure
Citation forms were elicited from participants using
pictures of objects presented on a computer screen
using Microsoft Powerpoint software. The pictures
presented single objects with which participants
were undoubtedly familiar, such as common animals,
furniture, types of vehicles, fruits, etc.
During the elicitation procedure, each participant
was seated opposite another signer of the same
language, and next to the computer used for the
presentation of pictures. Participants were instructed
to look at the computer screen and then to sign
to their interlocutors their sign(s) for the presented
object. During the entire procedure, the presentation
of pictures on the screen was controlled by a researcher, so that one picture was presented at a time,
and the next picture was presented after the sign
was produced clearly and to his/her satisfaction by
the participant. Since the younger ABSL participants
knew some ISL, they were explicitly asked to use
only their native (i.e., ABSL) signs. If a participant
produced an ISL sign, he or she was asked to sign
again, using the local sign.
The total number of elicitation pictures presented
to each signer was 35. Of the total number of elicited
signs, only 15 were analyzed for variation, for several
reasons. First, there were cases in which ABSL signers
did not seem to remember or know the sign for the
presented object, and instead produced sequences of
signs with related meanings, such as KNOBS FLAMES
OPEN-DOOR TRAY to refer to STOVE. Some of
these combinations were lexicalized as compounds.6
Lexicalized sequences often consisted of reduced
forms of individual signs, which were unsuitable
for a valid analysis of sublexical composition of the
individual signs. For example, reduction is known to
be related to stress pattern, and unstressed syllables
in multisyllabic signs are often displaced and become
temporally shorter (Sandler & Lillo-Martin, 2006).
We have not seen simplex words of more than one
syllable in the language, and we make the assumption that when monosyllabic signs are produced
in isolation, their surface forms more accurately
reflect their underlyingly form. We therefore chose
for the analysis only signs that were monosyllabic.7
A restriction on the choice of signs in the ASL data
was imposed by the fact that many responses were
fingerspelled.8 These were treated as phonologically
different and excluded from the analysis. Finally, in
6
7
8
4
5
Naturally, there are regional lexical differences in ISL, and
some ‘concepts’ have different signs representing them in
different areas. We attempted to avoid lexical variation by
choosing signers from the same area.
We are grateful to Carol Padden and Deniz Ilkbasaran for
collecting the ASL data used in this study.
Meir et al. (to appear) use four criteria to distinguish compounds from other sequences of signs. Compounds are
sequences that represent a single concept, are identical across
at least some signers and have at least two components
in common with other sequences produced for the same
referent. In addition, their production is natural and flowing
(as opposed to the hesitated signing which characterizes the
on-line construction of new sequences).
The notion of ‘syllable’ in sign language has been entertained
in several models and shown to be a real phonological
entity (for a comprehensive discussion of this issue see
Sandler & Lillo-Martin, 2006). The syllable is identified with
a single movement event (of any single type or simultaneous
combination of movement types – but not a sequence of
movements), analyzed as a syllable nucleus in some models
(e.g., Sandler, 1989; Perlmutter, 1993; Brentari, 1998).
Fingerspelling is the use of handshapes which represent letters
of the alphabet to spell a word borrowed from a spoken
language. It is a system which is phonologically distinct from
signing, although fingerspelled words may ultimately undergo
formational modifications to adhere to the phonological
constraints of the system of signs. Lexicalized fingerspelled
forms seem to be especially common in ASL.
each data set, and especially in ABSL, there were a
few sets of tokens that varied lexically. For example,
in ABSL, some participants signed HORSE with a
gesture representing the bit part of a bridle (a
handshape with the thumb and the index finger
pressed against each side of the signer’s face across
an open mouth); other ABSL signers signed HORSE
with an upside-down handshape “mounted” on
a
to represent a man sitting on a horse. Such
lexical variation found in a set of tokens rendered
it unsuitable for analysis of sublexical variation, and
signs that were lexically different were not included
in the comparison.
The list of 15 lexical items represented by the
data collected is given in Table 2. The first 11
items are shared by all three language sets, and
the remaining 4 items overlap only partly, because
of the constraints just mentioned. In any case, a
lexical match of the three language sets is of no real
methodological import, because of the arbitrariness
of the form-meaning relation in language. What
matters is sublexical variation for the same form in
a given language.
Table 2 – The lexical items represented in the collected data.
Items 1 – 11 were elicited from signers of all three languages and
are listed in the same central column. Since in each language four
of the items elicited were not shared by both other languages,
these items (12 – 15) are listed in a separate colum).
Item
number
ABSL; ISL; ASL
1
LEMON
2
SCORPION
3
TOMATO
4
CARROT
5
COW
6
DONKEY
7
FORK
8
LEAF
9
DOG
10
GOAT
11
TRAIN
ABSL
ISL
ASL
12
WOMAN
TELEVISION
WOMAN
13
BROOM
BROOM
EGG
14
STUFFED GRAPE LEAVES
FLOWER
FLOWER
15
GARLIC
CUCUMBER
CAT
21
3.3. Measuring variation
Naturally, variation cannot be compared across
different lexical items. It makes no sense to look at
the difference (i.e., variation) between surface realizations produced for different target representations.
If one signer represents whiskers for the concept ‘cat’
and another represents ‘cat’ as licking the paws, these
are two different lexical items, and not suitable for
comparison of sublexical variation. tThe essence of
variation is the existence of different variants of the
same item, which, in our case (i.e., at the sublexical
level), is a single lexical item. Therefore, for each
language, variatioisshould be measured first for each
of the 15 lexical items separately, and only then can
we combine these measures to get an indication of
the amount of variation at a more global level. This
methodology is outlined belo.. [WHAT DOES THIS
LAST SENTENCE MEAN ACTUALLY? HOW ELSE
WOULD IT BE MEASURED?]
For the analysis, we choose to use two measures
which we believe capture the essence of variation
in a way that is both transparent and simple. These
measures correspond to two important aspects of
variation: the range of the distribution and the extent
to which the data are concentrated or spread within
this range. In Statistics, when measuring the value
of a variable along a continuous scale (e.g., from
0 to 100), the range is the difference between the
highest and lowest observed values. Recall that in
the current analysis, for each token, the hand configuration component is specified in terms of discrete
phonetic features which cannot be considered in
terms of higher or lower values. This method of
coding is comparable to specifying the features [high]
[mid] and [low] for vowels, rather than measuring
their formant frequencies, since features are discrete
whereas frequency is measured along a continuous
scale. Therefore, for our purposes we may define
the range of variation in hand configuration as the
number of different features found across tokens.
To make things clear, let us consider a hypothetical
situation in which we have two different sets of ten
tokens for the sign FORK. For each set of tokens,
Figure 8 shows a distribution of features within the
subcategory Selected Fingers. We can see that in the
first set of tokens (represented by Distribution A)
there are two different finger selections: [I+M] ( ) and
[I+M+R] ( ). In the other set of tokens (represented
by Distribution B) there are three different finger
selections, [IMR], [IM] and [I] (as in ). In other words,
there are two different variants of SF in the first set
22
Figure 8 – Different amounts of variation indicated by the number of variants.
Distribution A
Category
Selected
Fingers
Signer
A
Signer
B
Signer
C
Signer
D
Signer
E
Signer
F
Signer
G
Signer
H
Signer
I
Signer
J
I+M
I+M
I+M+R
I+M
I+M
I+M+R
I+M
I+M
I+M
I+M
Signer
A
Signer
B
Signer
C
Signer
D
Signer
E
Signer
F
Signer
G
Signer
H
Signer
I
Signer
J
I+M
I+M
I+M
I+M
I+M+R
I+M
I+M
I
I+M
I+M
Distribution B:
Category
Selected
Fingers
of tokens and three variants of SF in the second
set. That is, the range of variation in SF is wider
in the second distribution. The number of different
features found within a set of tokens is therefore
an indicator of the range of a distribution. We shall
refer to this measure as the number of variants.
The other measure we will use provides information about the spread (or dispersion) of a distribution.
This is the number of tokens with the most frequent
feature. Consider the distributions in Figure 9. Since
in each distribution there are two different features
realizing the SF category, both distributions have the
same range of variation (number of variants = 2).
However, in Distribution A, eight out of ten tokens
have the same feature set [I+M], whereas in Distribution C the two feature sets are distributed more
equally across tokens: six tokens have the feature
[I+M] and four tokens have the feature [I+M+R].
We may say that in Distribution C tokens are more
“spread out” – i.e., they vary more – compared
to tokens in Distribution A. The frequency of the
modal feature – i.e., the number of tokens in which
the most frequent feature is found – is therefore an
important measure of variation. For convenience, we
shall refer to this measure as the mode (for further
discussion of the advantages and disadvantages of
this method, see Israel, 2009).
Beyond the single lexical item: a global measure
of variation
The ultimate aim of this study is to use the feature
by feature assessment of variation described above to
arrive at a more global measure of sublexical variation
within a language. The first step is to combine the
measures of each type calculated for all lexical items
by calculating the average mode value and average
number of variants for all lexical items within the
same language for a given phonological category,
as exemplified with hypothetical data in Table 3.
Recall that the first stage in the analysis was to
calculate the mode for each lexical item separately.
Figure 9 – Different amounts of variation idicated by the frequency of the modal feature.
Distribution A:
Category
Selected
Fingers
Signer
A
Signer
B
Signer
C
Signer
D
Signer
E
Signer
F
Signer
G
Signer
H
Signer
I
Signer
J
I+M
I+M
I+M+R
I+M
I+M
I+M+R
I+M
I+M
I+M
I+M
Signer
A
Signer
B
Signer
C
Signer
D
Signer
E
Signer
F
Signer
G
Signer
H
Signer
I
Signer
J
I+M
I+M
I+M
I+M+R
I+M+R
I+M
I+M
I+M+R
I+M
I+M+R
Distribution C:
Category
Selected
Fingers
23
Table 3 – Average mode and average number of variants as global measures of variation.
Category X
Item 1
Item 2
Item 3
Item 4
Item 5
Item 6
Item 7
Item 8
Average
Mode (%)
95
80
100
87
80
93
84
100
89.88 %
Number of
variants
1
2
2
1
1
2
1
3
1.63
In the above table, 95 represents the mode of the
set of tokens produced for Item 1, 80 was the mode
of all the tokens elicited for Item 2, etc. Once the
averages of all the modes and numbers of variants
have been calculated, we will already have reached
a more global representation of variation, since for
each phonological category we are left with two
measures per language: (1) average mode and (2)
average number of variants. Now it is possible to
compare the values of each of the two measures
of variation across languages. For example, if the
average mode calculated for Thumb Position is 90
for language A and 95 for language B, we may
say that with respect to this measure of variation,
language A shows more variation in thumb position
than language B. The two languages will also have
to be compared with respect to the average number
of variants.
Going one step up the phonological hierarchy, the
average measures calculated for subcategories may
be considered together in order to characterize the
degree of variation within each of the three major
categories. This, again, may be done by calculating
an average. That is, we will calculate an average from
the averages of each of the seven HC subcategories.
with the number of variants being highest in ABSL
and lowest in ASL. Within the Flexion subcategory,
the highest amount of variation was also found in
ABSL, but no difference was found between ISL and
ASL. The highest average number of SF variants was
the same in ABSL and ASL. The general ABSL > ISL
> ASL pattern was reversed for one category: Spread
Fingers. Altogether, ABSL showed more variation
than the other two languages on all measures, with
one exception, Spreading.
In order to check whether the differences found
are statistically significant, A Kruskal-Wallis test was
performed on the data. A highly significant difference
was found between the degree of variation in thumb
position in ABSL and those measured for ISL and
ASL. This was found for both the mode measure
(p<0.001) and the number of variants (p<0.01).
An instructive form of representation of the differences across ABSL, ISL and ASL is given in Figure
Figure 10 – Average mode values within Hand Configuration
subcategories.
4. Results
Using the methodology outlined in the previous
section, the study revealed a consistent ranking
across the three language groups with respect to the
amount of variation in hand configuration features
across tokens. As can be seen in Figure 10, in all
subcategories of this component, with the exception
of Finger Spreading, mode values were lowest in
the ABSL data and highest in the ASL data. That is,
as indicated by this measure and the data collected
for this study, for each subcomponent of Hand
Configuration, the amount of variation is greatest
in ABSL and ASL is the least variable.
This picture of the differences across the three
languages is made clearer by the second measure
of variation – the number of variants. Figure 11
shows that in four of the seven hand configuration
subcategories the same relative ranking was found,
Figure 11 – Average number of variants within each subcategory
of Hand Configuration.
24
Figure 12 – Ranges of average mode values.
Figure 14 – Examples of hand configuration variation in ABSL.
[Index]
[Index+Middle]
(a) STUFFED GRAPE LEAVES (variation in finger selection)
Figure 13 – Ranges of average number of variants.
[crossed]
[extended]
(b) WOMAN (variation in thumb position)
12 and Figure 13. This representation shows for
each language the range of average mode values
and the average number of variants presented in
Figure 10 and Figure 11 above.
The range measure clearly shows the differences
among the three languages with respect to both the
size of the range and its location along the Y-axis.
As for the mode, in ASL, its values are distributed
within the smallest range, about 17 percent. Like
ASL, ISL’s maximum value is 100 % (representing
a subcategory with zero variation), but its range
of average modes is wider – about 23%. Finally,
ABSL’s average modes spread over about 27 %.
Moreover, unlike ASL and ISL, none of the average
modes calculated for ABSL reached 100 %. In other
words, there was not a single subcategory of Hand
Configuration in which there was no variation across
ABSL signers. Figure 13 shows similar differences in
the range of average numbers of variants. Examples
of variation in ABSL are shown in Figure 14.
The analysis of variation shows that, in ABSL, ISL
and ASL, different tokens may vary along the sublexical features of hand configuration listed in Table 1.
Interestingly, when both measures of variation – the
mode and number of variants – are considered, a
robust pattern emerges: the amount of sublexical
variation is largest in ABSL and smallest in ASL, with
ISL in between the two. The fact that this pattern
was found in the majority of subparameters of hand
configuration suggests that these different amounts of
variation reflect some fundamental difference across
the three languages. This idea is further supported
by results from two additional analyses of variation
along features of location and movement – the two
other major parameters of sublexical form. Using
the same methodology, Israel (2009) found the
same cross-linguistic pattern of variation, namely
ABSL > ISL > ASL, for both Location and Movement.
It is therefore not only the Hand Configuration
component which varies to different extents, but
the entire form of lexical items.9
Assuming that all languages eventually develop
lexicons with highly conventionalized forms of items,
the results reported here suggest that ABSL, ISL and
ASL are currently situated at different points along
this conventionalization continuum. In the next
section we discuss the possible contributions of four
different factors to the development of regularity in
new languages.
9 In this study we did not include non-manual components,
such as facial expressions and mouthing, which in some
languages may be part of the lexicalized form.
5. Three different sociolinguistic backgrounds
We would like to put forth the hypothesis that, in
the early development of a sign language, an aggregate of sociolinguistic factors affects the convergence
of signers on a relatively fixed set of forms used
as lexical items. Underlying this hypothesis is the
assumption that convergence – i.e., transition from
more to less variation – is universal and characterizes all cases in which a new language emerges.
Intuitively, it does not seem possible for a new
language to exhibit the degree of regularity found
in well-established languages at the outset. In each
case of language emergence, however, the social
and linguistic settings, which obviously have an
impact on the way language develops, are unique.
Our hypothesis incorporates the following factors: a)
relation to other languages, b) the language’s age,
c) the size of the community, and d) the existence
of prescriptive norms. In this section we discuss
the possible influence of these factors and relate it
to the case at hand.
A new language may come to life in one of
two settings: within a community whose members
have no language at all, and within a community
whose members use different languages but none of
which is shared by all. The languages that emerge
in settings of the latter type are known as Pidgins
and, when passed on to children, as Creoles. Even
though these languages are fundamentally distinct
from any of the languages used natively by the
original members of the community, there is no
doubt that some grammatical elements are borrowed
from native languages into the pidgin/creole (e.g.,
McWhorter, 1997). This means that, compared to
a language developed by people who know no
language at all, pidgins and creoles get a head start.
The study of ISL and ASL has shown that both of
these languages developed in ways that resemble pidginization and creolization, with contributions from
German Sign Language, and other sign languages
of Europe, North Africa, and elsewhere in the case
of ISL (Meir & Sandler, 2008), and influence from
French Sign Language and local American varieties
in the case of ASL (Woodward, 1978; Fischer, 1996).
Therefore, in both cases, at the outset, there were
experienced signers who had been using signs
skillfully and consistently. That experience must
have been carried over into the new language, and
signers presumably did not lose their intuition about
the system which underlies sign production, even
25
when the forms themselves were new to them. In
contrast, ABSL has emerged in a relatively isolated
community and for many years was developed by
deaf people who had no knowledge of any other
language. It is reasonable to believe that for such
signers it takes longer to converge on a single form
for each concept.
If conventionalization is indeed gradual, then we
expect forms to be produced more consistently across
signers as the language gets older. A language’s age
may be measured not only in years but also in the
number of generations of users that have acquired
it. Young children have the capacity to acquire
and build upon the language as it is passed on to
them from a previous generation, as in the case of
Nicaraguan Sign Language mentioned above. It is
likely that children play an important role in the
process of convergence by taking the language a
step closer to fully conventionalized production of
lexical items. In our case, ASL is the oldest language,
which, according to our hypothesis, explains the
fact that it exhibits the lowest amount of sublexical
variation. However, since ABSL and ISL are of the
same age but vary to different extents, it is clear
that this factor by itself cannot predict differences
in the amount of variation.
The size of the community in which a language
develops may be another factor affecting the amount
of variation. Trudgill (1995) suggests that within
a small and isolated community there is likely to
be a large amount of shared information, and so
variation is more easily tolerated. This may well be
the case within the community of Al-Sayyid. When
much of the background information is shared by
interlocutors, it may be sufficient for a signer to
produce a token that approximates the overall image
which is conventionally associated with the target
concept in order for communication to succeed.
Metalinguistic awareness may have a strong impact
on language production. One aspect of such awareness is the idea that some forms are “better” or “more
appropriate” than others, and that certain forms are
“correct” and others are “incorrect”. Usually, these
concepts are shaped by authoritative sources, such
as schools, books, interpreters, and other influential
individuals, often associated with formality. On
this basis, it is reasonable to distinguish between
languages used in formal settings, such as ISL and
ASL, and languages whose users are not subject
to prescriptive pressure because it is never used
formally, such as ABSL. Thus, in both the ISL and
ASL communities there are Deaf organizations which
26
organize formal meetings and sign language courses;
members of the community meet in Deaf clubs; there
exist dictionaries of both languages; there are Deaf
theater and dance groups; and, finally, following
linguistic studies on both languages, the status of
both languages – mainly within, but also outside the
Deaf communities – has improved considerably. The
existence of prescriptive norms in ISL and ASL may
have affected the way signs are produced. In both
communities, dictionaries sign language instruction,
and interpreter training programs exist, which may
have the effect of establishing norms to some extent.
Such norms may in turn considerably reduce the
variety of alternate forms, thus pushing towards
more consistent signing. In the ASL community, the
normative sources just mentioned have longer histories and are therefore more established compared
to ISL, which could partly explain the differences
in the amount of variation found between the two.
In Al-Sayyid, where deaf people are fully integrated
into the larger hearing community, none of these
sociocultural developments has taken place, and,
to the best of our knowledge, the language is only
used informally.10
We propose that all of the sociolinguistic factors
just discussed played a role in the cross linguistic
differences found in this study. Table 4 shows that
each language has a different aggregate of these
factors. According to the discussion in this section,
the sum of factors is most conducive to convergence
in ASL and least conducive to convergence in ABSL.
The hypothesis developed above is motivated by
the amounts of variation measured in this study. In
order to test this hypothesis further, it is necessary
to measure variation in additional sign languages
with different aggregates of sociolinguistic factors.
We leave this investigation for future research.
Table 4 – A summary of cross-linguistic differences along sociolinguistic parameters.
ABSL
ISL
ASL
-
+
+
Age
~75
~75
~200
Population size
~150
~10,000
~500,000
-
+
+
Contribution from
other languages
Prescriptivism
10
A dictionary of ABSL signs is being compiled at the Sign
Language Research Lab in Haifa. At this point, however, the
dictionary is not available to ABSL signers.
6. Seeds of phonological form in ABSL
ABSL is a language by any functional measure.
Conversations about any topic relevant to the community take place in real time with no apparent effort
or hesitation. Humor is conveyed, stories are told.
There is a shared vocabulary. Even the variation that
we find in the lexicon is apparently well tolerated
if ease of communication is any indication, possibly
suggesting that the language simply has synonyms.
It is at once fascinating and surprising to discover
that a fully functional language appears to have fewer
conventionalized grammatical resources than other
more familiar and more established sign languages
(see Aronoff et al., 2008 for an overview). But a
closer look has revealed both a certain amount of
grammatical structure to support the system, as well
as the kernels of grammar in several areas. At the
syntactic level, though sentences are typically short,
strict word order within the clause is in place from
the beginning, showing that marking grammatical
relations between the predicate and its arguments
and between nouns and their modifiers is a fundamental ingredient – perhaps the most fundamental
ingredient – in human language (Sandler et al., 2005).
Other aspects of grammar are observably in the
process of becoming systematic. For example, productive ways of increasing the vocabulary are in
place, although much less regular in form than the
word order just described. Size and shape compounding or affixation is a widespread word formation
strategy, as is compounding in general (Meir et
al., to appear). In addition, a prosodic system that
demarcates constituents and signals dependencies
among clauses is developing, a system that is already
much more regular and systematic among young
adults of the second generation than among those
even twenty years older in the same generation
(Sandler et al., to appear).
When it comes to phonology, alongside the kind
of indeterminacy indicated by the present study,
we also see the seeds of a formal system taking
root. Evidence for the formation of phonological
categories will come, not only from minimal pairs
and sharp production of discrete forms, but also
from the participation of such units in processes
in the language which are related to form only,
and not to meaning. A typical example is assimilation. In established spoken and signed languages,
phonological assimilation is a regular process, in
which some set of units take on characteristics of
neighboring units under specific conditions. Nasal
assimilation in English involves the formal categories
nasals and stops, and the direction of assimilation
is regressive. All of these are formal properties,
unrelated to meaning.
Assimilation has been observed in ABSL under
certain conditions, spelled out in Sandler et al. (2009).
Although not yet general across the community,
these instances suggest the beginning of a formal
system. One example is found within a single family.
Because we have noticed that vocabulary is less
varied within families with several deaf people,
we have coined the term familylect to describe
this sociolinguistic entity. In one familylect, whose
members are a deaf mother and five deaf children,
we find assimilation in a lexical compound. The
compound CHICKEN^OVAL-OBJECT, meaning EGG,
is lexicalized across the village, but the assimilation
takes place only as signed by this family.
CHICKEN is produced with the index finger in
a curved shape, palm oriented downward, and
the hand bending at the wrist twice, apparently
motivated by the beak of a chicken pecking for
food. The sign for SMALL-OVAL-OBJECT is produced
with three spread fingers, the palm oriented up.
The hands for the basic compound are shown in
Figure 15a. In the familylect’s assimilated version,
the finger selection for the second sign assimilates
regressively to the first sign, CHICKEN. Figure
15b shows the hands for the sign, EGG signed
in a familylect. The deaf mother and all three of
her deaf daughters whom we recorded all signed
the ‘CHICKEN’ part of the compound with the
same assimilated form. Since they all signed it
the same way, we assume that the assimilated
form is lexicalized in this familylect. Crucially, the
assimilation is a purely formal process in which
the hand configuration of the second member of
the compound spreads to the first. Note that the
result is less iconic than the basic form, as chicken
beaks are pointed and not oval-shaped.
Alongside the presence of indeterminate phonological categories in this new language, we are beginning to see the buds of a system at the sublexical
Figure 15 – (a). Two handshapes for the compound CHICKEN^OVALOBJECT = EGG, standard form. (b). Consistent assimilation of handshape in EGG within a familylect.
27
level. This and other closely observed phenomena
show us how ABSL is moving toward phonological
organization.
7. Summary and conclusion
This study has shown that sign languages differ
in terms of the amount of variation in the form of
sign production exhibited across a community. The
amount of variation in the category of handshape
in a new language with little outside influence is
shown to be greater for nearly all subcategories of
that class than in languages with different social
histories. In particular, we find a cline of regularity
in form across ABSL, ISL, and ASL, such that ABSL
shows the most variation, ISL next, and ASL shows
the least amount of sublexical variation. These
and other related results (Israel, 2009; Sandler
et al., 2009) lead us to suggest first that, while
ABSL functions fully as a language and has certain
grammatical regularities, it has not yet developed
robust phonological categories. Differences in social
factors such as language age, size of a community,
and formal norms that hold between ABSL, ISL, and
ASL, are hypothesized to contribute to convergence
in language form.
Reference
1.
Ann, J. (1996). On the Relation between Ease of Articulation and Frequency
Occurrence of Handshapes in Two Sign Languages. Lingua, 98(1-3), 19-42.
2.
Aronoff, M., Meir, I., Padden, C. A., & Sandler, W. (2008). The Roots of
Linguistic Organization in a New Language. Interaction Studies, 9, 133-153.
3.
Barr, D. J. (2004). Establishing Conventional Communication Systems: Is
Common Knowledge Necessary? Cognitive Science: A Multidisciplinary
Journal, 28(6), 937-962.
4.
Battison, R. (1978). Lexical Borrowing in American Sign Language. Silver
Spring: Linstok Press.
5.
Battison, R., Markowicz, H., & Woodward, J. (1975). A Good Rule of Thumb:
Variable Phonology in American Sign Language. Analyzing variation in
language, 291-302.
6.
Brentari, D. (1998). A Prosodic Model of Sign Language Phonology. Cambridge, MA: MIT Press.
7.
Corina, D. (1993). To Branch or Not to Branch: Underspecification in Asl
Handshape Contours. In G. Coulter (Ed.), Current Issues in Asl Phonology
(Vol. 3, pp. 63-95). New York, San Francisco, London: Academic Press.
8.
Corina, D., & Sagey, E. (1989). Are Phonological Hierarchies Universal? Evidence
from American Sign Language. Paper presented at the ESCOL.
9.
Crasborn, O. (2001). Phonetic Implementation of Phonological Categories in
Sign Language of the Netherlands. Unpublished PhD dissertation, Leiden
University.
10. Crasborn, O., & van der Kooij, E. (1997). `Relative orientation in sign
language phonology’. In J. Coerts & H. de Hoop, eds., Linguistics in the
Netherlands 1997. John Benjamins, Amsterdam.
11. Fischer, S. D. (1996). By the Numbers: Language-Internal Evidence for
Creolization. International Review of Sign Linguistics, 1, 1-22.
28
12. Frishberg, N. (1975). Arbitrariness and Iconicity: Historical Change in
American Sign Language. Language, 51, 696-719.
13. Hockett, C. F. (1960). The Origins of Speech. Scientific American, 203, 89-96.
14. Hutchins, E., & Hazlehurst, B. (1995). How to Invent a Lexicon: The Development of Shared Symbols in Interaction. London: University College Press.
15. Israel, A. (2009). Sublexical Variation in Three Sign Languages. Unpublished
M.A. thesis, University of Haifa.
16. Kisch, S. (2000). Deaf Discourse: Social Construction of Deafness in a Bedouin
Community in the Negev. Unpublished MA thesis, Tel-Aviv University.
17. Liddell, S. K., & Johnson, R. E. (1989). American Sign Language: The
Phonological Base. Sign Language Studies, 64, 195-278.
18. Mandel, M. (1981). Phonotactics and Morphophonology in American Sign
Language. Unpublished PhD Dissertation, University of California, Berkeley.
19. McWhorter, J. H. (1997). Towards a New Model of Creole Genesis. New
York: Peter Lang.
20. Meir, I., Sandler, W., Padden, C., & Aronoff, M. (2009). Emerging sign
languages. In Oxford Handbook of Deaf Studies, Language, and Education,
Volume 2. M. Marschark and P. Spencer (Eds.).
21. Meir, I., Aronoff, M., Sandler, W., & Padden, C. (to appear). Sign Languages
and Compounding. In S. Scalise & I. Vogel (Eds.), Compounding: John
Benjamins.
22. Meir, I., & Sandler, W. (2008). A Language in Space: The Story of Israeli Sign
Language: Lawrence Erlbaum Associates.
23. Perlmutter, D. (1993). Sonority and Syllable Structure in American Sign
Language. In G. Coulter (Ed.), Phonetics and Phonology Vol. 3. Current Issues
in ASL Phonology. San Diego, CA: Academic Press.
24. Sandler, W. (1987). Sequentiality and Simultaneity in American Sign Language.
Unpublished PhD Dissertation, University of Texas, Austin.
25. --------- (1989). Phonological Representation of the Sign: Linearity and Nonlinearity in American Sign Language. Dordrecht: Foris.
26. --------- (1995). Markedness in the Handshapes of Signs: A Componential
Analysis. In J. v. d. Weijer & H. v. d. Hulst (Eds.), Leiden in Last: Holland
Institute of Linguistics Phonology Papers (pp. 369-399). The Hague: Holland
Academie Graphics.
27. --------- (1996). Representing Handshapes. International Review of Sign
Linguistics, 1, 115-158.
28. --------- (1999). Cliticization and Prosodic Words in a Sign Language. In T.
Hall & U. Kleinhenz (Eds.), Studies on the Phonological Word (pp. 223-254).
Amsterdam: Benjamins.
29. Sandler, W., Aronoff, M., Meir, I., & Padden, C. (2009). The Gradual Emergence
of Phonological Form in a New Language: Ms. University of Haifa, Stony
Brook University, and University of California San Diego.
30. Sandler, W., Meir, I., Dachkovsky, S., Aronoff, M., and Padden, C.. to appear.
The emergence of complexity in prosody and syntax. Shinichiro Ishihara
(ed.), special issue of Lingua on prosody and syntax.
31. Sandler, W., & Lillo-Martin, D. (2006). Sign Language and Linguistic Universals.
Cambridge, UK: Cambridge University Press.
32. Sandler, W., Meir, I., Padden, C., & Aronoff, M. (2005). The Emergence of
Grammar: Systematic Structure in a New Language. Proceedings of the
National Academy of Sciences, 102(7), 2661-2665.
33. Scott, D. A., Carmi, R., Elbedour, K., Duyk, G. M., Stone, E. M., & Sheffield,
V. C. (1995). Nonsyndromic Autosomal Recessive Deafness Is Linked to
the Dfnb1 Locus in a Large Inbred Bedouin Family from Israel. American
journal of human genetics, 57(4), 965.
34. Senghas, A. (2003). Intergenerational Influence and Ontogenetic Development in the Emergence of Spatial Grammar in Nicaraguan Sign Language.
Cognitive Development, 18(4), 511-531.
35. Senghas, A., Coppola, M., Newport, E., & Supalla, T. (1997). Argument
Structure in Nicaraguan Sign Language: The Emergence of Grammatical
Devices. In E. Hughes, M. Hughes & A. Greenhill (Eds.), The Proceedings of
the Boston University Conference on Language Development (Vol. 21, pp.
550-561). Boston: Cascadilla Press.
36. Trudgill, P. (1995). Dialect Typology: Isolation, Social Network and Phonological Structure. In G. R. Guy, C. Feagin, D. Schiffrin & J. Baugh (Eds.),
Towards a Social Science of Language: Papers in Honor of William Labov,
Volume 1: Variation and Change in Language and Society (pp. 3-22).
37. van der Hulst, H. (1993). Units in the Analysis of Signs. Phonology, 10(2),
209-242.
38. van der Kooij, E. (2002). Phonological Categories in Sign Language of the
Netherlands: The Role of Phonetic Implementation and Iconicity. Unpublished
PhD dissertation, Leiden University.
39. Woodward, J. (1978). Implicational Variation in American Sign Language:
Negative Incorporation. Sign Language Studies, 3, 20-30.
Imperativos Análogos a Raízes Infinitivas:
Evidência das Línguas de Sinais Americana e Brasileira
Imperatives function as analogues to root infinitives:
evidence from ASL and LBS
Diane Lillo-Martin*1,3 e Ronice Müller de Quadros**2,3
1
2
3
University of Connecticut
Universidade Federal de Santa Catarina
Haskins Laboratories
Resumo
Abstract
Neste artigo, investigamos a hipótese de Salustri e Hyams
(2003, 2006) de que (em algumas línguas) há imperativos
que funcionam como raízes infinitivas usados para expressar
significado de irrealis. As línguas em que isso acontece, não
apresentam o estágio de raízes infinitivas. Nós investigamos
esta hipótese a partir da produção de duas línguas de sinais
coletadas longitudinalmente em crianças surdas adquirindo a
língua de sinais americana (ASL) e a língua de sinais brasileira
(LSB). Nessas línguas, há dois tipos de verbos – um dos quais
prediz a realização de imperativos análogos. Os resultados
sustentam a hipótese do imperativo análogo de Salustri e
Hyams e, além disso, é mais uma evidência para a proposta
de Quadros (1999) baseada em dados de adultos. 
1. Contextualização
Frequentemente tem sido observado que crianças
muito pequenas produzem formas verbais não finitas
(com sintaxe apropriada), ao lado de formas corretamente flexionadas por volta dos dois anos de idade.
Poeppel & Wexler (1993) apresentam o exemplo em
(1) de Andreas, com 2,01. Em (1a), o verbo hab está
devidamente flexionado e na posição para verbos
flexionados, V2. Em (1b), o verbo haben está na
forma infinitiva e na posição para verbos infinitivos,
ou seja, no final da sentença (que pode aparecer
também em orações principais somente quando há
um verbo auxiliar na posição V2). Wexler refere
exemplos do tipo de (1b) como Infinitivos Opcionais
(OI), enquanto Rizzi utiliza o termo Raízes Infinitivas.
Neste artigo, usamos os dois termos indistintamente
para referir este tipo de exemplo.
In this paper, we investigate the hypothesis put forward by
Salustri and Hyams (2003, 2006), that (in some languages)
imperatives function as an analogue to root infinitives, being
used to express irrealis meanings in languages which typically
do not show a significant root infinitive stage.We investigate
this hypothesis by looking at longitudinal production data from
American Sign Language (ASL) and Brazilian Sign Language
(LSB), languages in which there are two verb types – only
one of which is predicted to behave like Italian in showing
an imperative analogue. Our results provide support for the
Imperative Analogue Hypothesis. In addition, our results provide
support for the analysis of sentences with these verb types
proposed by Quadros (1999) on the basis of adult data. 
(1) a. Ich hab ein dossen Ball.
I have a big ball
Eu tenho uma grande bola
b. du das haben
you that have
você que tem
Também foi observado que o uso dessas formas
varia significantemente entre as línguas (ver também
Guasti (1993/4), Phillips (1995), Wexler (1998), Buesa
Garcia (2007) e Grinstead (2008)). Wexler (1998)
concluiu que as línguas nas quais há a marcação do
estágio de infinitivos opcionais são línguas que não
apresentam sujeitos nulos. O autor sintetiza esta generalização em (2). Note que nas línguas com sujeitos
nulos há algumas ocorrências de raízes infinitivas,
mas as crianças aprendem que nessas línguas se usa
muito menos este tipo de construção, enquanto que
as crianças de línguas que não tem sujeitos nulos
usam raízes infinitivas com muita freqüência.
* [email protected]
30
(2) Generalização Sujeitos Nulos/Infinitivos Opcionais: A criança passará pelo estágio de OI se
somente se a língua não licencia sujeitos nulos.
No entanto, Salustri & Hyams (2003/2006) observaram que raízes infinitivas normalmente tem interpretação mood/irrealis e são eventivas1. Por isso e
por outras razões, elas argumentam que as raízes
infinitivas apresentam uma base gramatical e representam, portanto, uma manifestação universal. De
acordo com as autoras, existe um “núcleo universal”
do estágio de raízes infinitivas que se refere ao
fato de todas as crianças manifestarem a aquisição
de mood. O que é universal no estágio de raízes
infinitivas é o mapeamento do mood irrealis em uma
estrutura oracional sem marcação de tempo. Para
as crianças que estão aprendendo uma língua de
sujeito nulo marcado, uma estrutura sem marcação
de tempo análoga será usada para expressar o mood
irrealis. Nesse caso, a forma imperativa será usada
como análoga a raízes infinitivas.
Salustri & Hyams não estão dizendo que todas
as manifestações de imperativos apresentam todas
as interpretações encontradas para infinitivos. Por
exemplo, os imperativos não podem expressar
intenções futuras. No entanto, eles são irrealis e
eventivos e, nesse sentido, há um paralelo entre
as raízes infinitivas usadas pelas crianças que estão
aprendendo uma língua sem sujeitos nulos e o uso
de imperativos por crianças aprendendo uma língua
com sujeitos nulos. Para sustentar esta proposição,
Salustri & Hyams mostram que os imperativos são
muito mais frequentes na aquisição de línguas
com sujeito nulos do que em línguas sem sujeitos
nulos. Para evidenciar que esses resultados não
são determinados por questões culturais, as autoras
apresentam também dados de crianças bilíngues
adquirindo uma língua de cada tipo. As crianças
bilíngues do par italiano-alemão usam em torno
de 30 a 60% de imperativos no italiano enquanto
que no alemão, usam apenas em torno de 10%, no
período de 2;0 a 2;7.
2. Morfologia verbal na ASL e na LSB
Nós testamos a hipótese de imperativos análogos
de Salustri & Hyams observando a aquisição de
línguas que apresentam os dois tipos de verbos,
um que permite o licenciamento de sujeitos nulos
por meio da flexão de concordância e outro que
não licencia sujeitos nulos. Tanto a ASL como a LSB
tem verbos com concordância pessoal e locativa
que licenciam sujeitos nulos e verbos ‘simples’ sem
concordância que não licenciam sujeitos nulos (LilloMartin 1986; Quadros 1997).
LSB e ASL (assim como a maioria das línguas de
sinais) tem diferentes tipos de verbos que variam
entre verbos modificados para indicar argumentos
verbais e aqueles que não modificam. A descrição
dos verbos com concordância usados aqui é uma
modificação do que foi proposto por Meir (1998,
2002). Os verbos indicam seus argumentos pela
direção da face da mão e pelas posições iniciais e
finais dos sinais no espaço de sinalização. As locações
no espaço de sinalização podem ser associadas com
referentes de pessoas ou com locativos. Então, os
verbos de concordância pessoal (geralmente verbos
de transferência) tem a face da mão direcionada
para os seus objetos e movem de uma localização
associada com seu sujeito (+humano) para a localização associada com o seu objeto (+humano)2.
Os verbos simples não tem flexão para indicar o
sujeito e o objeto, mas podem ser sinalizados em
uma determinada localização indicando a locação do
evento, neste caso apresentando um comportamento
similar ao dos verbos de concordância locativa.
Esta flexão locativa e o movimento do verbo é
conhecida como concordância, ilustrada na Figura
1 (ver também Padden, 1988[1983]).
Figura 1 – Concordância verbal na ASL
a. I-ASK-HER
a. eu-pergunto-ela
Ambas flexões de concordância pessoal e locativa
licenciam argumentos nulos (Lillo-Martin 1986; Quadros 1997). Para a presente proposta, nós agrupamos
os verbos com concordância em oposição aos verbos
sem concordância (verbos ‘simples’).
2
1
Será mantida a palavra ‘mood’ no inglês. Uma tradução
para ‘mood’ no poderia ser ‘estado de ânimo’ associado
a ‘irrealis’ para expressar ‘intenções, desejos e vontades’.
b. SHE-ASKS-HIM
b. ela-pergunta-ele
Ver Quadros e Quer (2009) para uma problematização
da relação da trajetória da concordância com a noção de
transferência, além de não necessariamente termos ALVO
em verbos de concordância, mas sim TEMA.
31
Imperativos Análogos a Raízes Infinitivas
3. Primeira questão da pesquisa
3.1. Contextualização
Dada a existência de dois tipos de verbos na
ASL e na LSB, nossa primeira questão de pesquisa
é se o uso de imperativos será diferente entre as
sentenças com cada tipo de verbo, como sintetizado
em (3).
(3) Os imperativos são usados como análogos a
infinitivos com verbos concordância, mas não
com os verbos simples na ASL e na LSB?
A hipótese do imperativo análogo e a hipótese
da não existência de analogia dos imperativos apresentam diferentes previsões quanto a realização dos
imperativos nas línguas, baseado na suposição de
que não há diferença no uso de imperativos entre
as línguas, assim como apresentado em (4).
(4) Hipóteses
Hipótese dos imperativos análogos – imperativos
com verbos com concordância > imperativos com
verbos simples
Hipótese da não analogia dos imperativos – imperativos com verbos com concordância = imperativos
com verbos simples
3.2. Participantes
Nós analisamos o desenvolvimento da morfologia
verbal em duas crianças surdas adquirindo suas
línguas de sinais com os seus pais surdos: SAL,
adquirindo a ASL, entre 1;07 e 2;03 e de LEO,
adquirindo a LSB, entre 1;9 e 2;05. As crianças foram
filmadas longitudinalmente nteragindo espontaneamente com sinalizantes fluentes.3 O número de
enunciados analisados com um verbo é apresentado
na Tabela 1.
Tabela 1 – Número de enunciados analisados com um verbo
Idade
Leo
Sal
1;07
17
1;08
85
1;09
56
69
1;10
17
6
1;11
34
19
2;00
25
112
2;01
69
2;02
121
159
2;03
79
93
2;04
92
2;05
110
3.3. Método
Na ASL e na LSB, os verbos simples, normalmente, são de ‘estado’ e os verbos com concordância são ‘eventivos’. Como os imperativos são
geralmente eventivos, nós esperaríamos encontrar
mais imperativos com verbos com concordância
do que com verbos simples. Portanto, nós não
podemos simplesmente comparar o número de
imperativos com os verbos com concordância e os
verbos simples. Nós temos que considerar apenas os
verbos eventivos das duas categorias verbais. Assim,
os resultados reportados aqui distinguem o uso de
imperativos entre os verbos eventivos simples e os
de concordância.
Nós codificamos manualmente cada enunciado
declarativo com verbo eventivo, identificando-os
como eventivos ou de estado. Quando então tínhamos todos os verbos eventivos, classificamo-los
como imperativos e não-imperativos. O contexto e
o acento foram usados para identificar as formas
imperativas.
3.4. Resultados
3
Esses dados foram coletados como parte de um estudo
mais geral da aquisição da ASL e LSB, na Universidade de
Connecticut, Gallaudet University e Universidade Federal de
Santa Catarina em estudos envolvendo o campo da sintaxe.
Ver Chen, Lillo-Martin e Quadros (em preparação) para
explicações mais extensivas sobre coleta de dados e métodos
de transcrição.
Os resultados de nossas análises estão apresentadas na Figura 2.
Ambas as crianças produziram mais imperativos
com verbos com concordância do que com verbos
simples. Os resultados são mais surpreendentes porque
há uma produtividade muito maior de verbos simples
do que de verbos com concordância (ver Quadros
e Lillo-Martin, 2007). Seguem exemplos de uso de
imperativos na produção das crianças em (5) e (6).
Imperatives - Agr vs. Plain - Sal
32
0.50
0.40
Figura 2 – Resultados do primeiro estudo
0.30
Imperatives - Agr vs. Plain - Sal
0.50
0.20
Agr
Plain
0.40
0.10
0.30
0.00
Agr
1;7-1;8
1;9-1;10
1;11-2;0
2;2-2;3
1;7-1;8
1;9-1;10
1;11-2;0
2;2-2;3
Plain
0.20
0.10
0.00
Imperatives - Agr vs. Plain - Leo
.5000
.4000
.3000
Imperatives - Agr vs. Plain - Leo
.5000
.2000
Agr
Plain
.4000
.1000
.3000
-
Agr
1;8-1;9
1;10-1;11
2;1-2;2
2;3-2;5
Plain
.2000
(5).1000
Sal (1;09)
IX(bag) PICK-UP<imp>; IX(bag) PICK-UP<imp>
BAG. 1;8-1;9
1;10-1;11
2;1-2;2
2;3-2;5
HEY! BAG, PICK-UP<imp> IX(bag).
‘Pick up that bag; see that bag – pick it up! Hey!
Pick up the bag!’
Pega aquele pacote; veja aquele pacote – pega
ele! Pega o pacote!
(6) Leo (2;01)
PEGAR<imp>; PEGAR<imp> BALA; PEGAR<imp>
BALA IX<lá>; PEGAR<imp>; PEGAR<imp>.
‘Pega, pega aquela bala; pega a bala lá; pega, pega!’
Os resultados confirmam a hipótese do imperativo
análogo.
4. Segunda pergunta da pesquisa
4.1. Contextualização
Hoekstra & Hyams (1998), Deen & Hyams (2006),
e Salustri & Hyams (2006) fazem as seguintes observações em (7) e (8).
(7) Efeito de referência mood
Frequentemente as raízes infinitivas tem significado de um mood/irrealis.
(8) Hipótese de oposição semântica
A expressão do mood irrealis na gramática
inicial exclui a especificação temporal.
Estas observações levam Hyams e seus colegas a
postularem que várias formas não finitas são usadas
para expressar mood irrealis na gramática da criança.
Nós temos visto que em algumas línguas a forma
não finita usada é infinitiva, enquanto em outras é
o imperativo (e ainda há outras formas, tais como
subjuntivo ou outras formas não finitas).
Nós observamos no primeiro estudo que os
imperativos são usados pela SAL e pelo LEO para
expressar mood irrealis. Assim, podemos perguntar
se todas, ou virtualmente todas, as intenções irrealis
de SAL e LEO são expressas por meio da forma
imperativa. Se não, que formas são usadas? Está
questão é a base da formulação da segunda pergunta
da pesquisa em (9):
(9) Como o mood irrealis é expresso na ASL e na
LSB da criança?
Nós percebemos que os adultos sinalizantes da ASL
e da LSB utilizam uma marca não manual que acompanha as expressões de desejo e intenção (irrealis). Esta
marca não é usada para expressar o futuro simples.
É muito difícil julgar se as crianças estão usando
essa marca não manual, pois o ângulo da filmagem
normalmente não permite acesso a esta informação
e também por outras dificuldades de visualizar suas
faces. Todavia, fica claro que elas não estão usando
consistentemente a marca não manual de irrealis.
4.2. Participantes e metodologia
Foram analisadas algumas sessões do desenvolvimento da SAL e do LEO assim como examinado no
primeiro estudo. Foram codificados os enunciados
eventivos com um verbo de mood: realis vs. irrealis,
separando-se os verbos com concordância dos
verbos simples.
4.3. Resultados
Os resultados destas análises estão apresentados
na Figura 3.
Vários aspectos são identificados a partir destes
dados. A primeira coisa é que tanto SAL como LEO
praticamente não produzem verbos irrealis além
de imperativos. Depois, ambos produzem os mood
irrealis imperativos e não imperativos como os
dois tipos de verbos, com concordância e simples.
33
Figure 3 – Resultados do segundo estudo
Mood Development – Agr Verbs - Sal
Mood Development – Plain Verbs - Sal
1.00
1.00
0.80
0.80
0.60
0.60
Realis
Irrealis
Imperative
0.40
0.40
0.20
0.00
0.20
1;7
1;8
1;9
1;11
2;0
2;2
0.00
2;3
Mood Development – Plain Verbs - Leo
1.00
0.80
0.80
Realis
Irrealis
Imperative
0.60
0.40
1;7
1;8
1;9
1;11
2;0
2;2
2;3
Mood Development – Agr Verbs - Leo
1.00
Realis
Irrealis
Imperative
0.60
0.40
0.20
0.20
0.00
Realis
Irrealis
Imperative
0.00
1;8 1;9 1;10 1;11 2;1 2;2 2;3 2;4 2;5
LEO, como já identificado anteriormente, não utiliza
imperativos com verbos simples.
Ao analisar os verbos com concordância, fica claro
que ambas as crianças usam formas imperativas e
não imperativas para marcar irrealis. Então, não
podemos afirmar que o imperativo ‘análogo’ é a
forma exclusiva de realizar o mood irrealis nas
produções de SAL e LEO.
As formas irrealis não imperativas geralmente servem para descrever eventos futuros ou possibilidades,
frequentemente com a marcação do verbo PODER
ou NÃO-PODER. Veja alguns exemplos em (10).
(10) a. Sal 1;09
RING-LOC(here) TABLE; IX(FAT) COMELOC(here) HERE.
‘(He will) bring the table here; he (will) come
here.’
Ele trará a mesa aqui; ele virá aqui.
b. Sal 2;02
WRITE-LOC(easel) CAN.
‘(You) can write on the easel.’
Você pode escrever no quadro.
SAL também fez uso de um emblema g(wait-aminute) para acompanhar as formas não imperativas
de irrealis usadas por ela. Alguns exemplos são
dados em (11). Este emblema é uma forma gestual
convencionalizada usada culturalmente por ouvintes
1;8 1;9 1;10 1;11 2;1 2;2 2;3 2;4 2;5
americanos, por isso chamamos de emblema ao
invés de sinal.
(11) a. Sal 1;09
COME-LOC(here) g(wait-a-minute).
‘(He will) come here in a minute.’
Ele virá aqui em um minute.
b. Sal 2;02
IX(self) HAT BRING-LOC(here) g(wait-aminute)
‘I will bring the hat here in a minute.’
Eu trarei o chapéu aqui em um minuto.
Não somente as formas não imperativas de irrealis
produzidas por SAL são acompanhadas pelo emblema
g(wait-a-minute), mas muitas outras formas. Talvez
ela esteja usando este emblema como uma forma
de marcação de irrealis.
Morford e Goldin-Meadow (1997) verificaram
em crianças utilizando sinais caseiros este mesmo
gesto convencional (glosado por elas como WAIT)
com os mesmos sentidos. As autoras pontuaram
que “In addition to using the gesture for this conventional meaning, the deaf children also used the
gesture to identify their intentions, that is, to signal
the immediate future” (Morford & Goldin-Meadow
1997, p. 429)4.
4
“Além de usar o gesto com o seu sentido convencional, as
crianças surdas também o utilizaram para identificar suas
intenções, ou seja, para indicar o futuro imediato”.
34
Morford e Goldin-Meadow verificaram que os
pais não sinalizantes das crianças usaram o mesmo
gesto no sentido convencional, que foi caracterizado
como ‘pedido’ de tempo ou de ‘espera’. mas elas
observaram também que os pais não usaram o
gesto para identificar o futuro imediato, como as
crianças fizeram.
A mãe da SAL utiliza o gesto g(wait-a-minute), mas
não conseguimos analisar o seu uso. Não fica claro
se o uso deste gesto na ASL já está gramaticalizado,
mas nos dados da criança essa possibilidade pode
ser considerada.
O que acontece com os verbos simples? Fica
evidenciado que as crianças não utilizam formas não
imperativas irrealis com verbos simples (embora SAL
produza algumas formas imperativas com os verbos
simples). Estas formas parecem ser infinitivas – não
marcada temporalmente – de acordo com a Hipótese
da Oposição Semântica. Na verdade, é difícil afirmar
se essas formas dos verbos simples são ‘infinitivas’,
uma vez que os verbos simples não necessariamente
apresentam marca de tempo. Muitos, apesar de
não serem todos, verbos simples da produção de
SAL estava acompanhado de g(wait-a-minute); por
outro lado, nós não podemos afirmar se as crianças
usam formas não marcadas de tempo para expressar
irrealis com verbos simples.
5. Estrutura
Salustri e Hyams levantam uma questão quanto
a razão de em algumas línguas os infinitivos serem
usados para expressar irrealis, enquanto em outras
as formas imperativas (ou outras) serem utilizadas.
As autoras propõem que a derivação das formas
infinitivas é mais econômica do que a derivação de
imperativos, porque o imperativo envolve movimento
do verbo para projeções mais altas de Mood e Force.
Portanto, o infinitivo sempre será a forma usada a
não ser que seja bloqueado por alguma razão. Em
italiano, elas propõem que o infinitivo tem que
checar traços abstratos de Agr. Uma vez que o verbo
se move para Agr, T é necessariamente checado e
o infinitivo não será usado para expressar mood
irrealis. Neste caso, portanto, o imperativo é usado
e não a forma infinitiva.
Como isso pode ser aplicado na derivação da
diferença entre os verbos simples e com concordância quanto ao uso de imperativos na ASL e na
LSB? Quadros (1999) propôs que os verbos simples
e com concordância projetam estruturas diferenciadas na LSB (Quadros, Lillo-Martin & Chen Pichler,
2004 estendem esta proposta para a ASL). Quadros
apresenta evidências de que a projeção de Agr é
necessária para os verbos com concordância, mas
não para as sentenças com verbos simples. Nessa
análise, os verbos simples não requerem checagem
de traços abstratos de concordância, mas sim há um
afixo de Tempo que se combina com o verbo por
meio da afixação (affix hopping). A estrutura em
(12) é proposta para os verbos simples com irrealis.
(12) Estrutura para verbos simples com irrealis
MoodP
M
[+irrealis]
VP
Não há necessidade de usar uma estrutura mais
complexa incluindo Força para expressar irrealis,
uma vez que os verbos simples podem ser combinados com irrealis sem uma projeção mais alta.
Por outro lado, os verbos com concordância
requerem o movimento para as projeções de Agr.
Na proposta de Quadros, Agr e T são projetados
com verbos de concordância. A estrutura em (13)
é necessária para os verbos com concordância com
irrealis. Então, uma vez que há necessidade de
projeções mais altas de qualquer forma, seguindo
a proposta lógica de Salustri e Hyams, irrealis pode
ser expresso usando uma forma imperativa.
(13) Estrutura para verbos com concordância
com irrrealis (imperativos)
ForceP
Force
[+imp]
AgrP
Agr
MoodP
M
[+irrealis]
VP
6. Conclusão
Nós verificamos que as crianças adquirindo ASL
e LSB usam imperativos com muito mais freqüência
com verbos com concordância do que com verbos
simples.
Isso é consistente com a Hipótese do Imperativo Análogo, se nós considerarmos os verbos com
concordância como línguas do tipo do italiano e os
verbos simples como línguas do tipo do alemão.
No entanto, nós verificamos que as criançãs não
tem problemas com o uso de formas imperativas
irrealis (tanto com verbos simples como com verbos
com concordância). Essas formas incluem expressão
de futuro e mood e, talvez, algumas delas sejam
associadas à irrealis com uma forma específica
associada para significar isso: g(wait-a-minute).
Os resultados de nosso estudo oferecem elementos
para corroborar a proposta de Quadros (1999) de
que os verbos simples e os verbos com concordância
apresentam estruturas diferenciadas; em particular,
de que os verbos simples não envolvem movimento
para projeções de Agr, mas os verbos concordância
sim. Então, este estudo é mais um exemplo de que
é possível estabelecer uma relação entre as análises
gramaticais e os estudos da aquisição da linguagem.
Referências
1.
Buesa Garcia, Carlos (2007). Root infinitives are compatible with nullsubject languages: evidence from Spanish. Presented at GALA, Barcelona.
2.
Deen, Kamil Ud & Hyams, Nina (2006). The morphosyntax of mood in early
grammar with special reference to Swahili. First Language 26.1, 67-102.
3.
Hoekstra, Teun, and Hyams, Nina (1998). “Aspects of root infinitives”.
Lingua 106:91-112.
4.
Grinstead, John & Pratt, Amy (2007). The optional infinitive stage in
child Spanish. In A. Belikova, L. Meroni & M. Umeda (Eds.), Proceedings of
Generative Approaches to Language Acquisition – North America 2, 351-362.
Cascadilla Press.
5.
Guasti, Maria Teresa (1993/4). Verb syntax in Italian child grammar: finite
and nonfinite verbs. Language Acquisition 3, 1–40.
6.
Lillo-Martin, Diane (1986). Two Kinds of Null Arguments in American Sign
Language. Natural Language and Linguistic Theory 4, 415-444.
7.
Meir, Irit (1998). Thematic structure and verb agreement in Israeli Sign
Language. PhD dissertation, Hebrew University of Jerusalem.
8.
Meir, Irit (2002). A cross-modality perspective on verb agreement. Natural
Language and Linguistic Theory 20.2, 413-450.
9.
Morford, Jill & Goldin-Meadow, Susan (1997). From here and now to
there and then: The development of displaced reference in Homesign
and English. Child Development, 68.3, 420-435.
10. Padden, Carol (1988 [1983]). Interaction of Morphology and Syntax in
American Sign Language: Outstanding Dissertations in Linguistics. New
York: Garland. (Originally distributed as: PhD dissertation, University of
California, San Diego.)
11. Phillips, Colin (1995). Syntax at age two: Cross-linguistic differences. In:
C. Schütze, J. Ganger & K. Broihier (eds), Papers on Language Processing
and Acquisition. MITWPL #26, 225-282.
35
12. Poeppel, David & Wexler, Ken (1993). The Full Competence Hypothesis
of Clause Structure in Early German. Language 69, 1-33.
13. Quadros, Ronice Müller de (1997). Aspectos da sintaxe e da aquisição da
língua de sinais brasileira. Letras de Hoje, 110, 125-146.
14. Quadros, Ronice Müller de (1999). Phrase Structure of Brazilian Sign Language. Unpublished Ph.D. Dissertation, Pontifícia Universidade Católica
do Rio Grande do Sul, Porto Alegre, Brazil.
15. Quadros, Ronice Müller de, Lillo-Martin, Diane, & Chen Pichler, Deborah
(2004). Clause structure in ASL and LSB. 26th Annual Conference of the
German Linguistics Association (DGfS); Mainz, Germany.
16. Quadros, Ronice Müller de & Lillo-Martin, Diane (2007). Gesture and the
acquisition of verb agreement in sign languages. Proceedings of the 31st
Annual BUCLD, 520-531. Cascadilla Press.
17. Quadros, R. M. de e Quer, J. (2009). A caracterização da concordância nas
línguas de sinais. In Sintaxe na língua de sinais brasileira. Organização de
Heloisa Salles. Editora Unb. (no prelo).
18. Salustri, Manola & Hyams, Nina (2003). Is there an analogue to the RI
stage in the null subject languages? Proceedings of the 27th Annual BUCLD,
692–703. Cascadilla Press.
19. Salustri, Manola & Hyams, Nina (2006). Looking for the universal core of
the RI stage. In Torrens, V. & Escobar, L. The Acquisition of Syntax in Romance
Languages. Amsterdam: John Benjamins.
20. Wexler, Ken (1998). Very early parameter setting and the unique checking
constraint: A new explanation of the optional infinitive stage. Lingua,
106, 23–79.
Sign production by first-time hearing signers:
A closer look at handshape accuracy
Produção de gestos por ouvintes gestuantes iniciados:
um olhar atento para a orientação manual
Deborah Chen Pichler*
Gallaudet University
Abstract
Resumo
This paper presents phonetic analysis of hand configurations
elicited from hearing adults exposed for the first time to signs in
American Sign Language. The accuracy of their production is
analyzed in terms of various handshape sub-features, including
degree of finger splay and opposition of the thumb. Two familiar
factors from spoken second language acquisition, markedness
and phonological transfer, are proposed as plausible factors
affecting subjects’ handshape accuracy. Although these
conclusions are preliminary, based only on a limited data sample,
they indicate promising directions for further study of hearing
adults learning a sign language as a second language. Research
attention in this area stands to greatly deaf children and their
parents, the vast majority of whom are hearing and need to
become proficient in sign language as efficiently as possible.
Este artigo apresenta a crítica das configurações manuais
elicitadas de adultos ouvintes expostas pela primeira vez à
Língua Gestual. O rigor de sua produção pode ser analizado em
termos das subestruturas que suportam a orientação manual,
incluindo a abertura dos dedos e a oposição do polegar. Dois
factores familiares na aquisição de línguas orais, enquanto L2 – a
marcaçãoe a transferência fonológica são factores plausíveis de
afectação do rigor da orientação manual nos sujeitos. Apesar de
estas conclusões serem preliminares e baseadas, apenas, em
dados de amostragem, elas indicam difecções prometedoras
para a estudo de adultos ouvintes que tentem a língua gestual
como sua L2. A investigação nesta área é necessária tendo
em conta a maioria de pais ouvintes com filhos surdos cuja
necessidade de proficiência em Língua Gestudal é mandatória.
Keywords: sign language; language development; deafness;
linguistics 
Palavras Chave: Língua Gestual – Desenvoltura de
Linguagem, Surdez, Linguística 
Introduction
This paper discusses methodological issues related to research on the acquisition of natural sign
language1. Although sign acquisition has received
much research attention over the past half century,
nearly all of that research has focused on very
young deaf and hearing children or (less frequently)
Natural sign languages such as American Sign Language
(ASL) or Língua Gestual Portuguesa (LGP) have evolved
over time within deaf communities and have rich lexical
and grammatical structures independent from the spoken
languages that surround them. They are not to be confused
with artificial sign systems such as Signed Exact English (SEE)
or Signed Portuguese. The latter systems were invented by
educators for the purpose of teaching spoken language and
are not naturally transmitted from deaf parent to deaf child
in the way that natural sign languages are.
* [email protected]
1
deaf adults who were not permitted to learn sign
language until puberty or beyond, leading to late
acquisition of their first language. The study discussed
in the current paper is unusual in that focuses on
an entirely new case of language learning, that of
hearing adults learning a natural sign language as
a second language. This aspect of sign language
acquisition has been severely understudied, despite
its clear potential for improving deaf children’s
chances for normal linguistic development.
At first glance, it seems improbable that studies of
hearing adults learning sign language as a second
language should have any relevance for linguistic
development of deaf children. Extensive research
on the transmission of natural sign languages within
deaf, signing families has demonstrated repeatedly
that when deaf children are exposed to sufficiently
rich input in a natural sign language, they acquire
38
a first sign language in a manner that is comparable
to first language acquisition of speech by hearing
children. The question remains, however, as to how
native-like sign language input must be in order to
qualify as “sufficiently rich.” The vast majority of
deaf children (about 95% in the US, according to
Mitchell et al. 2006) are born to hearing parents who
are unlikely to know a natural sign language such
ASL or Lengua Gestual Portuguesa (LGP). Therein
lies the value of research on sign acquisition as a
second language. Although some of these parents
commit to learning sign language, aware of the
enormous benefits of sign language exposure to
early linguistic development and later academic
achievement (Mayberry and Eichen 1991; Wilbur
2005), they have limited pedagogical resources at
their disposal, which hinders their progress. As
first-time adult learners of a sign language, these
parents have to cope not only with difficulties typical
of second language acquisition in general (e.g.
transfer or interference from the first language), but
also difficulties specific to learning language in a
new modality (manual/visual versus oral/aural). A
better understanding of how second-language and
second-modality (M2 for short) factors interact is
key to maximizing the speed with which hearing
parents can become proficient in natural language
and provide their deaf children with the early input
critical to their linguistic development.
This paper discusses results of a small pilot project
focused on one very narrow aspect of M2 acquisition
of sign language, handshape accuracy. For readers
unfamiliar with sign linguistics, a Background section
is provided, providing basic information about sign
phonological structure and summarizing previous
work on phonological acquisition in sign language
first and second language acquisition. This paper
also addresses a fundamental methodological issue
raised by these previous studies: the importance
of a consistent phonetic notation system for sign
languages.
Background
Signs can be described in terms of four basic
phonetic parameters: hand configuration (sometimes
referred to as handshape2), location, movement,
2
Although handshape and hand configuration have often
been used interchangeably, I follow Johnson (2008) and
his colleagues in distinguishing between the two terms. I
will use hand configuration to refer to actual instances of
production by a given individual, while handshape will refer
Image 1 – ASL sign for MOTHER
and palm orientation. For example, the ASL sign for
MOTHER in Image 1 consists of repeated contact
(movement) at the chin (location) of the 5 handshape3
oriented with the palm facing the contralateral side of
the signer’s body. While small variations in parameter
values are common across different signers, forms
that deviate too far from the standard in one or
more parameters are judged by native signers to be
ill-formed and may not be understood. For instance,
shifting the location of the ASL sign MOTHER up
a few centimetres to the upper lip renders the sign
meaningless. Accurate production of the phonetic
parameters of signs is thus an important goal for
new learners of sign language.
Markedness and transfer
For new signers, much like for new learners of
spoken languages, a variety of factors may prevent
accurate phonetic production. One is L1 (first language) transfer, or the tendency of substituting a
phonetic form in the learner’s existing L1 phonetic
inventory for a phonetic form from the new language,
due to the mistaken perception that the two are
interchangeable. Another factor affecting accurate
phonetic production for second language learners
is markedness, or variability in the relative ease
with which certain phonetic units are recognized
and reproduced. Unmarked forms generally occur
more frequently, are more easily perceived and are
easier to articulate than marked forms.
The concepts of markedness and transfer are familiar from studies of second language acquisition of
spoken languages (Jakobson 1968), but they are also
3
to the abstract label (based on the ASL manual alphabet and
number system) traditionally used to group similar-looking
hand configurations into convenient categories. Thus an
individual asked to reproduce a list of signs targeting the Y
handshape may produce multiple distinct hand configurations.
For a list of common ASL handshapes, the reader is directed
to Image 2, appearing later in this paper.
Sign production by first-time hearing signers: A closer look at handshape accuracy
applicable to L1 and M2 acquisition of sign languages.
Boyes Braem (1973, 1990) made the earliest proposal
for an anatomically-informed hierarchy of handshape
markedness for ASL. She observed that the fingers of
the hand are successively bound by ligaments into
several autonomous and semi-autonomous bundles,
such that certain combinations of fingers are more
difficult to manipulate than others. The thumb and
index are the most independent of digits, and as such
are easy to manipulate individually. In contrast, the
ring, middle and pinky fingers are bound together
by a ligament, making them difficult to manipulate
independently. From these anatomical observations,
Boyes Braem predicted that handshapes involving
independent manipulation of the thumb or index (eg.
the 1 handshape) are less marked than handshapes
requiring individual manipulation of the remaining
fingers (eg. the 3 handshape).
In addition to purely anatomical factors for markedness, Boyes Braem proposed secondary factors that
potentially increase articulatory complexity, such as
crossing or insertion of fingers (for handshapes R or
T and N, respectively), or opposition of the thumb
(eg. in handshapes S and 1). Taken together, these
factors predict a hierarchy of handshape complexity
that Boyes Braem divided into 4 stages (plus A as
the maximally unmarked handshape), listed below.
Image 2 – Boyes Braem (1973/1990) hierarchy of handshape
markedness
ASL handshape font courtesy of http://www.lapiakdesign.com/lapiakasl.html
The predictions of the Boyes Braem hierarchy have
been tested on naturalistic output by young children
acquiring ASL as their L1 and have generally been
found to be consistent with patterns of handshape
acquisition and substitution: unmarked handshapes
are not only produced earlier than marked handshapes, they are also commonly substituted for the latter
(Boyes Braem 1990 and McIntire 1977). Given these
effects in L1 signing, it is reasonable to suppose
39
that markedness may exert similar effects on M2
signing. Two recent studies of M2 sign phonology,
Rosen (2004) and Chen Pichler (to appear), differed
in their portrayal of markedness as a factor in M2
handshape accuracy. Rosen (2004) maintained that
adult learners have fully developed motor skills and
are less unlikely to struggle with the anatomical
demands of marked handshapes as young children
do. However, Chen Pichler (to appear) argued that
a fully developed motor system does not guarantee
flawless motor skills in new physical domains (also
noted by Mirus et al. 2001). Thus adult M2 learners
may be susceptible to markedness effects, particularly
in the beginning of their sign language development.
Unlike markedness, which is often assumed
to apply universally, regardless of the learner’s L1
and L2, phonological transfer is a language-specific
phenomenon. Transfer is said to be positive when
a learner correctly perceives a target form as being
identical to a form in his or her L1. The learner can
then transfer that form into the L2 without having
to learn it. On the other hand, transfer is said to
be negative when the learner fails to perceive a
difference between the target form in the L2 and
a similar (but not identical) form in the L1. In this
case, the learner fails to establish a new phonetic
category for the new form, instead assimilating it to
an existing form in the L1. Thus persistent foreign
accent, perhaps contrary to intuition, stems more
from mispronunciation of target forms that are highly
similar to familiar forms in L1, than from forms that
are completely foreign to the learner (Best 1995;
Flege 1995).
Once a phonetic form is recognized by the learner
as new, its acquisition is predicted to take place
according to the normal developmental path, subject
to universal developmental factors such as markedness (Major 2001). Chen Pichler (to appear) argued,
contra Rosen (2004), that transfer also exerts an effect
on M2 sign acquisition, despite disparate modalities
of the L1 and L2. This is because the source for
phonological transfer in these cases lies not in the
spoken L1, but in the system of conventionalized
gestures (e.g. emblems) commonly used in hearing
communities. Image 3 shows two conventionalized
gestures that are widely used and understood across
the American hearing community.
Although they do not qualify as formal phonetic
analyses of the handshapes employed by conventionalized gestures, popular “guides to American
gesture” such as Axtell (1997) and Armstrong and
Wagner (2003) employ a set of plain-English han-
40
Image 3 – American conventionalized gestures “call me” (on the
left) and “wait a minute” (on the right)
dshape labels in their descriptions, such as fist or
point or V-shape. This practice suggests that such
handshapes are reasonably similar and identifiable
across users. Many of these handshapes look identical
or similar to those found in ASL, raising the possibility that new ASL learners might transfer them to
their M2 signing4. For instance, returning to Image
3 presented earlier, the “call me” gesture utilizes a
hand configuration that is potentially conflatable with
the ASL Y handshape. Although Y is categorized
as marked under the Boyes Braem hierarchy, new
signers confronted with this handshape in ASL signs
may nonetheless reproduce it accurately, due to
the fact that they have experience using it in the
conventional gesture “call me.” In this way, proposes
Chen Pichler (to appear), markedness and transfer
may exert opposing effects on the second language
development of handshape.
Importance of a phonetic notation system for sign
Although studies of handshape development are
relatively common in the sign language literature,
particularly with respect to first language development, comparability between these studies is
hindered by the lack of a standardized way to notate
details of sign form. Unlike spoken languages, for
which linguists can represent phonetic forms with
considerable detail using the Intenational Phonetic
Alphabet (IPA), there is currently no widely available notational system to represent the phonetic
distinctions of signed forms. Most studies (including
Chen Pichler, to appear) use the traditional labels
introduced by Stokoe et al. (1965) in the first ASL
4
This possibility is based on the assumption that nonsigners
recognize handshape as a discrete subunit of gestures at
some level.
dictionary to refer to whole handshapes. These labels
are based on the manual alphabet and number
system in ASL (e.g. Y, F, 1, 3), with additional labels
for configurations that do not correspond to any
letter or number (e.g. open-8 or b(aby)-O). While
referring to “the Y handshape” or “the 1 handshape”
is convenient for informal discussion, such global
designations are grossly inadequate as a substitute
for a phonetic notation system. Small but potentially contrastive differences go unrepresented and
potentially distinct hand configurations are lumped
together in a single category. For instance, although
the two forms in Image 4 differ noticeably in the
degree of abduction of the thumb and pinky, both
are designated as the Y handshape.
Image 4 – Two variations of the Y handshape
ca l l m e
WRONG
We still know little about the distribution of these
two configurations in ASL or other sign languages,
or the extent to which they are interchangeable. Yet
we cannot begin to address such questions using
notational systems that conflate the two forms.
Sign notation systems that represent signs by
specifying the values for each of their four parameters rather than a global designation are a marked
improvements over the traditional handshape labeling
system, and have been available since Stokoe et
al. (1965). Notation systems proposed by (Prillwitz
et al. 1989) and Liddell and Johnson (1989) can
capture a far greater level of phonetic detail than
previous systems and have been adopted by more
recent studies of sign phonological acquisition (eg.
Takkinen 2002). The latest and most comprehensive
notation system to be proposed, Johnson (2008),
allows sign forms to be represented in overwhelming
detail. While such a system may strike one as overly
exhaustive in its descriptive power, it is precisely
this feature that permits researchers the means to
finally determine what level of phonetic detail is
contrastive in natural sign languages and important
for language acquisition.
Material and Methods
Goals of the study
There are two goals of the current study. The main
goal is to look for effects of handshape markedness
and handshape transfer in the phonological production of first-time M2 signers. This goal is expressed
by two null hypotheses: (a) subjects production of
unmarked handshapes will be more accurate than
their production of marked handshapes, and (b)
subjects will substitute (transfer) a handshape from
conventional gesture for a target sign handshape
whenever the subject’s gestural rendition of that
handshape is identical to the target sign handshape.
In cases where the subject’s gestural rendition of the
handshape is in fact not similar but not identical to
the target sign handshape, such transfer will result
in an error (negative transfer).
The second goal of the current study is to compare
the patterns that emerge from a traditional wholehandshape label system compared to highly detailed
phonetic notation system. The majority of the data,
previously reported in Chen Pichler (to appear), were
coded using the traditional global-labeling system
plus minor modifications to specify features such as
thumb opposition and finger abduction. The current
report includes a comparison of the previously
reported analysis with a new analysis based on
partial recoding of the data using the Johnson (2008)
notation system, probing the extent to which coding
practices influence the generalizations that emerge
in data analysis.
Subjects and stimuli
Subjects for this pilot project were four hearing,
nonsigning adults (two male and two female) with
no previous experience learning a sign language.
These adults were not technically ASL learners,
since they were not enrolled in an ASL class. Age of
exposure to a second language, the environment in
which it is learned, the type and amount of exposure
the learner receives, the attitude and motivation he/
she brings to the task, etc. all affect development.
The net result of these factors is that even within
the same classroom, individual students can progress
at vastly different rates. By testing subjects with no
previous experience with sign languages, this study
aimed to mimic the very initial stage of acquisition,
before learner variability becomes too pronounced.
Also, since this experiment constituted subjects’
first and only input to ASL, this allowed a higher
41
degree of control over subjects’ exposure to the
target language than would have been possible if
they had been enrolled in an ASL class.
Stimuli included five common American gestures
and 16 signs from ASL, representing the handshapes
S, 1, B-dot, Y, W, and open-8 (the full list of stimuli
is given in Appendix A at the end of this report).
According to the Boyes Braem hierarchy, both
marked handshapes (2/5 gestures and 7/16 signs)
and unmarked handshapes (3/5 gestures and 9/16
signs) were represented. Among the stimuli were
also signs with handshapes potentially identical or
very similar to handshapes used in conventional
gesture (14/16 signs), and signs with handshapes
distinct from any used in gestures (2/16 signs).
To minimize confounding effects of phonological
complexity in the other parameters, only signs with
relatively unmarked location (either chest, chin
or neutral space) and movement (mostly simple
contact movements or a shake/trill) were selected,
and no more than a single handshape throughout
the sign was allowed (i.e. signs with sequences of
multiple handshapes were excluded). The resulting
16 ASL signs and 5 conventionalized gestures were
presented in random order, in two trials. The signs
were modeled by a native signer from a deaf, signing
family, filmed from two angles to give as clear a
view of each sign as possible. Subjects saw each
stimulus three times: first from head-on, then from
the side, and finally from head-on again, before
being given three seconds to copy the stimulus.
Subjects were instructed to try to reproduce each
item as faithfully as they could, focusing on the
hands.
All subject production was videotaped and coded
for handshape accuracy. A reproduced sign hand
configuration was coded as accurate if it was the
same as that used by the model in terms of handshape category (identified by the traditional global
labeling system) and any of the following features that were applicable: thumb opposition (fully
opposed, partially opposed or unopposed) position
of unselected fingers (open or closed), degree of
splay (adduction or abduction) of extended fingers
(hyper splayed, fully splayed, splayed or unsplayed),
and relative position of thumb and pinky when
in contact (pinky pinned under thumb, pinky and
thumb tips pressed against each other, thumb resting
against side of pinky). These criteria were also used
for coding subjects’ gestures, although these were
not categorized as accurate or inaccurate, on the
assumption that subjects would likely produce their
42
customary forms for familiar gestures rather than
faithfully imitating the model.
The production of subject 4 was subsequently
recoded in more detail, using notation from the
Johnson (2008) system to specify multiple features
concerning extension/flexion of the thumb and
each finger, adduction/abduction of the four fingers,
thumb opposition and abduction, and thumb-finger
contact. Appendix B at the end of this report lists the
possible values for each feature and their notational
symbols as proposed by Johnson (2008).
Results
Image 5 below presents a visual summary of
subject’s percent accuracy in reproducing the six
target handshapes under discussion in this chapter.
For this initial analysis, a token could only be counted
as accurate or inaccurate; tokens in which some
aspect of the produced hand configuration matched
the target, but others did not, were classified as
inaccurate. In other words, hand configurations were
evaluated as whole units, following typical practice
in previous studies of handshape acquisition.
The handshapes in Image 5 are grouped visually
by shading pattern: unmarked handshapes (S, 1,
B-dot) are represented in solid shading, while marked
handshapes (Y, W, open-8) appear in patterned shading (checkered or striped). An absent bar represents
a 0% accuracy rate for that particular handshape.
Evaluating accuracy on the level of whole hand
configurations, subject 4 scored very high for target
Image 5 – Table 2: Handshape accuracy in elicited signs
handshapes Y and W; her production matched
that of the target 100% of the time (8/8 and 4/4,
respectively) for finger/thumb splay and thumb
opposition. In contrast, subject 4 scored quite low
for target handshapes S (1/8), 1 (2/6), B-dot (2/10)
and open-8 (1/4).
Subject 9 performed at high accuracy for target
handshapes 1 (100% or 6/6) and B-dot (90% or 9/10).
He was moderately accurate for target handshapes
S (63% or 5/8), Y (63% or 5/8) and open-8 (50% or
2/4). All of his production of target W was coded
as inaccurate (0% or 0/4) because his pinky was
consistently pinned beneath his thumb, a configuration that contrasted with the target forms.
Subject 10 was highly accurate in his production of
the target S handshape (100% or 8/8) and moderately
accurate in his production of target 1 (67% or 4/6)
and B-dot (60% or 6/10). He was less accurate for
target handshape Y (25% or 2/8), due to the fact
that his pinky and thumb were splayed further apart
than the model for these signs. His production for
both the open-8 handshape and the W handshape
were all coded as inaccurate (0% or 0/4 in each
case). Like subject 9, his main error in production
of target W involved the relative placement of the
thumb and pinky.
Finally, subject 12 was moderately accurate in her
production of five out of the six target handshapes:
S (50% or 4/8), 1 (83% or 5/6), B-dot (70% or 7/10),
Y (75% or 6/8) and open-8 (50% or 2/4). The only
handshape she did not produce accurately was
target W (0% or 0/4), for largely the same reason
43
Image 6 – Errors with target open-8 handshape: FEEL
Target
Subject 4
Subject 10
Subject 12
Errors with target W handshape: SIX-YEARS-OLD
Target
Subject 9
as subjects 9 and 10 (inaccurate relative placement
of the thumb and pinky, or contact between the
thumb and ring finger).
Discussion
Handshape accuracy for elicited signs: Effects of
markedness
The prediction for this study with respect to
markedness was that all else being equal, subjects
would reproduce unmarked handshapes more
accurately than marked handshapes. Applied at
the level of overall performance for each subject,
this prediction appears to be true only for subject
9, whose production was overall more accurate for
unmarked handshapes than for marked handshapes.
This was not the case for any of the other subjects,
least of all for subject 4, whose performance was
overall more accurate for marked than unmarked
handshapes.
However, examination of the data with respect to
specific handshapes rather than pooling overall performance reveals plentiful evidence that markedness
should still be considered as an influential factor
affecting M2 handshape accuracy. For instance,
all subjects but subject 4 were able to reproduce
unmarked 1 and B-dot handshapes 60% or more of
the time. Conversely, highly marked handshapes W
and open-8 posed serious problems for all subjects.
With the exception of subject 4, none of the subjects
Subject 10
Subject 12
correctly reproduced the W handshape in any of
the sign stimuli (0/4 items). As for open-8, subjects
9 and 12 reproduced the handshape accurately in
half of the sign stimuli (2/4 items), subject 4 in a
quarter of the stimuli (1/4 items) and subject 10 in
none of the stimuli (0/4 items). The examples below
illustrate the errors with target handshapes open-8
and W, respectively. For each example, the target
stimulus viewed by the subjects appears furthest
to the left.
The open-8 handshape is predicted to be very
marked by the Boyes Braem hierarchy and to my
knowledge does not occur in any conventionalized American gesture. It is thus unfamiliar to the
subjects and predicted to be acquired according
to normal developmental patterns, i.e. subject to
universal factors such as markedness. Errors with
this handshape generally involved placement of the
nonselected fingers (thumb, index, ring finger and
pinky). Errors with the W handshape, in contrast,
were due to relative placement of the pinky and
thumb. For instance, subjects 9, 10 and 12 produced
instances of this handshape with the nail of the pinky
pinned under the thumb, rather than touching padto-pad with the thumb. Although this appears to be
a handshape variant permitted in ASL, it was coded
as an error, since it did not match the handshape
in the stimulus that these subjects received. In the
case of subject 10, errors with target W may also
be explained by transfer, to be discussed in the
next subsection.
44
Subjects also made many substitutions of a less
marked handshape for one that is more marked.
For instance, subjects 4 and 9 substituted the A
handshape for target S in the sign SENATE. S is
considered by Boyes Braem (1973/1990) to be more
marked than A because it requires opposition of
the thumb, a secondary feature she claims increases
handshape complexity. Since these substitutions
are made in the direction of less marked forms,
such errors are still compatible with the prediction
that markedness exerts a negative influence on
accuracy. In the case of subject 4, her particular
pattern of substitution could alternatively be analyzed
as the result of transfer, as I will detail in the next
subsection.
Finally, subject 4 was highly successful with
marked handshapes W and Y, reproducing them
accurately in 100% of the sign stimuli. This result
is unexpected, from the viewpoint of the Boyes
Braem (1973/1990) hierarchy. One feature that the
W and Y handshapes have in common, in contrast
to the open-8 handshape, which subject 4 reproduced poorly, is that the index, middle and ring
fingers shared the same configuration (i.e. all open
or all closed). These three fingers are not bound
together in the same way that, for instance, the
middle, ring and pinky fingers are. Thus a strictly
anatomical/production explanation fails to account
for this particular accuracy pattern. Instead, the
movement of the inside three fingers as a single
block, in opposition to the pinky and thumb, may
have improved the perceptual saliency of W and
Y for subject 4, leading to successful reproduction.
Handshape accuracy for elicited signs: Effects of
transfer
The second prediction for this pilot study was that
subjects would accurately reproduce handshapes
that are identical to a handshape they use for a
conventional gesture (positive transfer), but commit transfer errors for target handshape that are
very similar but not identical to a handshape they
use for conventional gesture (negative transfer). In
the original Chen Pichler (to appear) analysis, two
handshapes were categorized as “very similar but
not identical” if they differed only in one of the
features used to determine accuracy of handshape
reproduction. Of the handshapes analyzed here,
only open-8 has no similar gestural counterpart.
The other five handshapes are all similar or identical
to handshapes found among common American
Image 7 – Subject 4 errors in S handshape
Target WORK
Target SENATE
Target SYMBOL
Subject 4 WORK
Subject 4 SENATE
Subject 4 SYMBOL
gestures (see Appendix B) and are thus potential
sources for transfer.
Transfer can be posited in cases where subjects
produced a nontarget sign handshape that matched
a handshape they also used in gesture. Our data
include several such cases of such negative transfer,
almost all involving unmarked handshapes S and
1. Subject 4 substituted a fist with unopposed or
partially opposed thumbs for the target S handshape
(fully opposed thumb) in several signs (WORK,
SENATE, SYMBOL), as illustrated in the examples
below.
As mentioned in the previous sub-section,
markedness might be a factor in this substitution
pattern, since the S handshape is considered to
be slightly more marked than the A handshape.
However, markedness alone does not provide a
satisfying explanation for these errors. Both A and
S handshapes lie at the unmarked extreme of the
markedness hierarchy, which should render both of
them relatively easy to execute, especially for adults
Image 8 – Subject 4’s production of the S handshape in gesture
Target gesture Yes!
Subject 4 gesture Yes! (trial 1)
accurate reproduction (at least with respect to thumb
position), as seen in the sign ATHLETE below. For
signs targeting the S handshape, transfer results in
an error, despite the highly unmarked status of the
S handshape.
A negative transfer account is also plausible for
errors in subject 4’s production of the target handshape 1.The sign stimulus WHERE requires full
thumb opposition, but was reproduced by subject 4
in both trials with an unopposed thumb. Subsequent
analysis reveals that this subject’s gesture for wait
a minute was also produced with the unopposed
thumb version of the 1 handshape (as illustrated
below). The same variant of the 1 handshape occurred in subject 4’s production of the sign DIFFERENT,
but only on the dominant hand.
Image 10 – Subject 4’s production of target handshape 1
Subject 4 gesture Yes! (trial 2)
with fully developed motor abilities. More likely,
some other factor has led subject 4 to perceive
the S handshape incorrectly as the A handshape in
these stimuli.
Close examination of subject 4’s production of the
gesture Yes! (two fists raised in the air in victory)
provides a possible explanation for her handshape
errors described above. While the model produced this gesture with two S handshapes, subject 4
reproduced it with unopposed or partially opposed
thumbs, as illustrated below. This makes a compelling
case for negative transfer, since the same nontarget
handshape occurs across both gesture and sign
stimuli targeting the S handshape.
I propose that subject 4’s handshape inventory
includes a handshape that we can call the fist handshape. When she sees signs or gestures with the
A and S handshapes, she perceptually assimilates
them (Best 1995) to the fist category she already
possesses. This leads her to reproduce signs targeting the A and S handshapes with her particular
version of the fist handshape, which involves an
unopposed or partially opposed thumb. For signs
targeting the A handshape, this transfer results in
Image 9 – Subject 4’s production of the A handshape
Target ATHLETE
45
Subject 4 ATHLETE
Target WHERE
Target Wait a minute
Subject 4
WHERE (trial 1)
Subsject 4
WHERE (trial 2)
Subject 4 Wait a minute
In the initial analysis of the data, almost all instances of negative transfer occurred with unmarked
handshapes. The only exception was the case of
W, already mentioned in the previous subsection.
This highly marked handshape (or something like
it) exists in conventional American gesture, namely
in the gesture for three (not to be confused with
the ASL sign THREE). In the stimuli, our signing
model demonstrated this gesture with the tip of her
pinky finger pinned under the pad of her thumb.
In contrast, for both sign stimuli using the W handshape, our signing model did not use her thumb
to pin down her pinky finger, but either placed it
beside the thumb (WATER) or pressed its pad against
the pad of the thumb (SIX-YEARS-OLD). Subject
46
Image 11 – Sign model and subject 10 handshapes for three,
WATER, SIX-YEARS-OLD
Image 12 – Subject 10 handshape for call me and WRONG
Target call me
Target three
Target WRONG
Subject 10 three
Subject 10 call me
Target WATER
Subject 10 WATER
Target SIX-YRS-OLD
Subject 10 SIX-YRS-OLD
10 pinned his pinky finger under the thumb for
both instances of the gesture three as well as for
all four sign tokens calling for the W handshape.
This pattern suggests that subject 10 may have
perceptually assimilated the model’s W handshapes
to an existing handshape category in which the tip
of the pinky is pinned under the thumb, resulting in
negative transfer in his sign production. Comparison
of subject 10’s handshapes with those of the sign
model are shown below.
Finally, negative transfer from gesture did not
occur in all cases where it was predicted to occur.
For instance, the target Y handshape in the sign
stimulus WRONG differs from the handshape in the
target gesture call me in the degree to which the
thumb and pinky are splayed. The two handshapes
are otherwise very similar, and one might have
expected subjects to transfer their handshape from
the gesture call me to signs requiring the Y handshape. However, this did not occur in the majority
of cases. As mentioned earlier, all four subjects
accurately reproduced the gesture call me, with the
Subject 10 WRONG
pinky and thumb splayed widely. They also (even
subject 10) accurately reproduced the Y handshape
of the target sign WRONG, in which the pinky and
thumb were not widely splayed.
In this case I agree with Rosen (2004) that marked
handshapes do not necessarily pose the same production challenges for M2 learners as they do for L1 child
learners whose motor skills are still developing. After
all, even complete ASL novices, such as the subjects
in this study, were able to accurately reproduce the
ASL handshape in WRONG after seeing it for the
first time. Nevertheless, comparison of these cases of
non-transfer with the cases of transfer documented
earlier for subject 4 suggests that markedness may
actually exert a subtle influence in both cases. While
subject 4 was able to perceptually extract unmarked
handshapes 1 and S from sign stimuli and “recognize”
them as being part of her existing gestural inventory,
highly marked handshape Y apparently did not
trigger this same kind of recognition for subjects,
and thus was not subject to transfer. Alternatively,
of course, subjects’ success in producing distinct
Y configurations for their gestures and signs could
indicate that they accurately distinguish the two
in their perception. The data sample for this pilot
study is too limited to rule out either interpretation.
Results from recoding of Subject 4 data
Subsequent to the initial data analysis reported
above, the data from subject 4 was recoded using
the recent sign phonetic notation system recently
presented by Johnson (2008). This exercise was
intended as a small test of the claim, often made by
Johnson and others, that researchers can see very
different patterns in the data depending on what
kind of notational system they adopt for coding. The
effects of recoding on the interpretation of the data
from subject 4 can be represented by the specific
examples summarized below.
(1) Recoded data led to the same basic generalizations as the initial coding, but greatly
increased the degree of certainty with which
these generalizations could be made. The initial
analysis of subject 4’s production of the 1 handshape
identified an error pattern by which the thumb of her
dominant hand was consistently in an unopposed
position, leading to the proposal of transfer as a factor
affecting her production accuracy. Recoding did not
change this result, but it did significantly facilitate
the coding process. According to Johnson (2008), the
feature I initially coded as thumb opposition depends
on the joint effects of palmar abduction (the degree
to which the thumb is extended forward away from
the palm) and radial abduction (the degree to which
the thumb is extended laterally away from the index
finger). Coding each of these sub-features in turn
was significantly more straightforward than coding
for thumb opposition, with the result that the second
analysis proceeded more quickly and with far less
tortuous equivocation than the initial analysis.
(2) Recoded data revealed generalizations that
were previously overlooked due to lack of detail.
The initial coding focused on the degree to which
the pinky and thumb were splayed (abduction),
roughly determined by the angle between these
digits and the nearest neighboring finger. Despite
this relative straightforwardness of coding, it did not
lead to a clear overall pattern of hand configuration
accuracy. In particular, several of the configurations
coded as having splayed pinky and thumb (like the
target form) nevertheless still looked different from
the target form.
Recoding of these hand configurations revealed
that this difference lay in the degree of flexion of
the closed fingers: in the target forms of WRONG,
SAME and MEASURE, the first set of knuckles (MCP)
are only partially flexed, while they are fully flexed
in the gesture call me. As a result, the full length
of the metacarpal bone (c) is visible to the camera
in WRONG, and fingers 2, 3 and 4 appear loosely
closed in SAME and MEASURE. In both subject 4’s
gesture and sign production, the MCP was fully
flexed, altering the overall appearance of the hand
configuration. Although this difference was immediately noticeable, the initial analysis could not
capture it, since it did not consider finger flexion.
47
The reanalysis revealed that while the model used
distinct configurations for gestures and signs, subject
4 used the same configuration for both, representing
another potential example of negative transfer that
was missed in the original analysis.
(3) Recoded data led to a reinterpretation of
the data that directly contradicted the interpretation under the initial coding method. Although
there were no cases in which initial coding and
recoding of the same feature resulted in clearly contradictory results, I did encounter a related situation in
which the recoding process exposed a generalization
presented in the initial analysis that was inconsistent
with the stated coding criteria. According to the
initial analysis, none of the subjects scored above
50% accuracy for the open-8 handshape, a pattern
I attributed to the marked status of that handshape.
Recoding did not change the overall accuracy scores
for the subjects, but it revealed that subject error lay
mainly in the degree of flexion of the non-selected
digits (in this case, the thumb, index, ring finger and
pinky). The initial coding, which did not include
flexion, should have been limited to finger and
thumb splay, features that alone do not account
for the subjects’ low accuracy scores. Apparently,
my summary of the initial analysis was influenced
by subjective impressions about finger and thumb
position that were not formally included in the
coding process. Although the ultimate interpretation
on subject accuracy for open-8 remained unchanged
across both analyses, my error is a reminder that clear
coding criteria, based on an objective notation system
with sufficient phonetic detail, play an important
role in constraining researchers’ perception of the
data to only the features on which they claim to
base their analysis.
Conclusions
Previous research on the production of ASL signs
by new hearing adult learners of sign language
(M2 learners) discounts markedness and transfer as
potential factors affecting handshape accuracy (Rosen
2004). However, the limited pilot data presented here
provide support for individual effects of markedness
and transfer, both individually and interactionally,
in M2 signing. These preliminary data also suggest
that transfer (both positive and negative) is in some
cases blocked for highly marked handshapes. In the
terminology of speech perception models such that
advanced by Best (1995), markedness appears to be
48
a factor that can prevent learners from perceptually
assimilating certain handshapes to similar handshapes
that they use in gesture and for which they already
have an established handshape category. I propose
that in such cases, subjects approach the target sign
as an unfamiliar bundle of handshape, movement
and location features that they must do their best
to replicate in a short period of time. Their adult
cognitive skills are sufficient to ensure accurate
reproduction in some cases (e.g. the Y handshape
in WRONG for subject 10) but not in others (eg.
the W handshape in WATER for subject 4), where
they make errors reminiscent of those observed in
the L1 ASL of young signers.
Of course, markedness and transfer alone cannot
account for all the handshape errors that M2 signers
produce. In some cases marked handshapes were
reproduced with higher accuracy than expected,
even when these handshapes were distinct from
handshapes used in common American gestures,
and therefore assumed to be novel for our subjects.
In these cases, I agree with Rosen (2004) that the
cognitive abilities of adult learners sometimes prevail
over markedness, allowing for accurate reproduction of the target handshape where a child learner
might typically fail. This serves as a reminder of
the fundamental complexity of second language
acquisition in general: each adult learner brings a
unique combination of linguistic experience, aptitude
and motivation to the task of a new language, such
that no two learners will follow the same developmental path. When the new language also happens
to be of a different modality than the learner’s
native language, additional challenges may arise.
The most effective approach to M2 sign phonology
must recognize a variety of factors that influence
accuracy, as well as the complex ways in which
these factors may interact.
Bibliografia
1.
Armstrong, N. and M. Wagner (2003) Field guide to gestures: how to
identify and interpret virtually every gesture known to man. Philadelphia:
Quirk.
2.
Axtell, R. (1997) Gestures: The Do’s and taboos of body language around
the world. Wiley.
3.
Best, C. (1995) A direct realist vew of cross-language speech perception.
In W. Strange (ed.), Speech perception and linguistic experience: issues in
cross-language research. Baltimore, MD: York Press. 171-204.
4.
Boyes-Braem, P. (1973/1990) Acquisition of handshape in American
Sign Language. In V. Volterra & C. Erting (Eds.), From gesture to sign
language in hearing and deaf children, Springer-Verlag, 107–127.
5.
Chen Pichler, D. (to appear) Sources of handshape error in first-time
signers of ASL. In D. J. Napoli and G. Mathur (eds). Deaf around the
world. Oxford: Oxford University Press
6.
Flege, J. (1995) Second language learning: Theiry, findings, and problems. In W. Strange (ed.) Speech perception and linguistic experience:
issues in cross-language research. Baltimore, MD: York Press. 233-237.
7.
Jakobson, R. 1968. Child Language, Aphasia and Phonological Universals.
The Hague: Mouton.
8.
Johnson, R. E. (2008). Conversations toward an international phonetic
notation for signed languages. Handout from The Second Workshop on
the Phonetic Notation of Signed Languages, Université de Paris-Sud,
Orsay, France. June 2008.
9.
Liddell, S. and B. Johnson (1989). American Sign Language: the phonological base. Sign Language Studies, 64, 195-277.
10. Major, R. (2001) Foreign accent: The ontogeny and phylogeny of second
language phonology. Mahwah, NJ: Lawrence Erlbaum Associates.
11. Mayberry, R. I. and E. B. Eichen (1991). The long-lasting advantage of
learning sign language in childhood: Another look at the critical period
for language acquisition. Journal of Memory and Language, 30, 486-512.
12. McIntire, M. (1977) The acquisition of American Sign Language hand
con – figurations. Sign Language Studies 16:247-266.
13. Mirus, G., C. Rathmann, and R. Meier. (2001) Proximalization and
distalization of sign movement in adult learners. In V. Dively et al.
(eds.) Signed languages: Discoveries from international research, 103-119.
Washington, DC: Gallaudet University Press.
14. Mitchell, R., T. Young, B. Bachleda and M. Karchmer (2006) How many
people use ASL in the United States? Why estimates need updating.
Sign Language Studies 6(3): 306-38.
15. S. Prillwitz, R. Leven, H. Zienert, T. Hanke, J. Henning, et al. (1989) Hamburg Notation System for Sign Languages – An Introductory Guide.
International Studies on Sign Language and the Communication of the
Deaf, Volume 5., Institute of German Sign Language and Communication
of the Deaf, University of Hamburg.
16. Rosen, R. (2004) Beginning L2 production errors in ASL lexical phonology. Sign Language and Linguistics 7:31-61.
17. Stokoe, W.C., D.C. Casterline and C.G. Croneberg (1965). A dictionary
of American Sign Language on linguistics principles. Silver Spring, MD:
Linstok Press.
18. Takkinen, R. (2002). The secrets of handshapes: The acquisition of handshapes by native signers at the age of two to seven years. Translated
summary (154-172). Deaf Studies in Finland, Volume 1. Publisher:
Kuurojen Liitrrory. [Original title: Kasimuotojen salat: Viittomakielisten
lasten kasimuotojen omaksuminen 2-7 vuoden iassa.]
19. Wilbur, R. (2004). After 40 years of sign language research, what do
we know? In S. Bradaric-Joncic, & Ivasovic, V (eds.), Sign language, deaf
culture & bilingual education. Zagreb: University of Zagreb.
49
Appendix A: Stimuli list
Note: For two-handed signs in which both hands form the same handshape, each hand was counted
separately to arrive at the calculations of accuracy presented in Image 5.
Highly unmarked handshapes
S
Gesture: Yes!
Signs: WORK, SENATE, SYMBOL
1
Gesture: One/wait a minute
Signs: DIFFERENT, WHERE
Moderately unmarked handshapes
B-dot
Gesture: Stop!
Signs: MINE, YOURS, SCHOOL, PLEASE
Moderately marked handshapes
Y
Gesture: Call me
Signs: SAME, MEASURE, WRONG
W
Gesture: Three
Signs: WATER, 6-YEARS-OLD
Highly marked handshapes
open-8
Gesture: none
Signs: MEDECINE, FEEL
50
Appendix B: Handshape notational symbols from Johnson (2008)
Domain
Sub-domain
Thumb config
Th-Fing contact
Right (or Left) Hand
Configuration
Finger config
Feature tier
ext HC
CM Opp.
CM Abd.
MCP Flex.
DIP Flex.
Thumb Surf.; Bone
+/ – th/f contact
F Surf.; Bone; Nos.
1 MCP Flex.;
1 PIP Flex.
1 DIP Flex.
Abd./Cross 1-2
2 MCP Flex.
2 PIP Flex.
2 DIP Flex.
Abd./Cross 2-3
3 MCP Flex.
3 PIP Flex.
3 DIP Flex.
Abd./Cross 3-4
4 MCP Flex.
4 PIP Flex.
4 DIP Flex.
Categories of Flexion/ Extension
F: fully flexed
f: partially flexed
e: partially extended
E: fully extended
h: partially hyper-extended
H: fully hyper-extended
Categories of Abduction and Crossing
〈 widely abducted
< neutral (slightly abducted)
= adducted and adjacent
x more-ulnar crossed over more-radial, normal position
xp more-ulnar crossed over more-radial, more-ulnar on
tip of more-radial
xa more-ulnar crossed over more-radial, tip of more-ulnar
on bone a of more-radial
X
hyper-crossed
r more-radial crossed over more-ulnar, normal position
rb more-radial crossed over more-ulnar, tip of more-radial
of bone a of more-ulnar
CM Rotation:
O Opposed
U Unopposed (Neutral)
L Laterally aligned
CM Abduction
〈 widely abducted
< neutral (slightly abducted)
= adducted and adjacent
Possible Values
+, O, U, L;
〈, <, =
E,e,F,f
E,e,F,f
a,d,p,r,u; DI
+,a,d,f,r,u; D,I,P,M; 1,2,3,4
E,e,F,f
E,e,F,f
E,e,F,f
〈, <, =, x, xp, xa, X r, rp, ra
E,e,F,f
E,e,F,f
E,e,F,f
〈, <, =, x, xp, xa, X r, rp, ra
E,e,F,f
E,e,F,f
E,e,F,f
〈, <, =, x, xp, xa, X r, rp, ra
E,e,F,f
E,e,F,f
E,e,F,f
Oral language and sign language:
possible approaches for deaf people’s language development
Lingua orale e lingua dei segni:
approcci possibili per lo sviluppo del linguaggio nei sordi
Carmela Bertone*1 and Francesca Volpato**1
1
Dipartimento di Scienze del Linguaggio Università Ca’ Foscari di Venezia
Abstract
Abstract
Deafness is a sensory impairment which strongly affects
the normal acquisition and development of linguistic abilities.
Deaf people are severely hindered in the development of oral
speech because they do not have direct access to the linguistic
input and many of them do not acquire much more than
the rudiments of oral communication. While hearing children
acquire easily and naturally a spoken language, deaf children
might acquire in the same way a sign language, exploiting the
visual modality. This study investigated the general linguistic
competence in Italian of four different groups of deaf individuals
(orally-trained children with cochlear implants, native signers,
non-native signers and deaf foreigners adolescents and adults),
by using a standardized picture matching task, in order to
determine the level of their linguistic competence. Results
revealed that most deaf individuals showed a performance
comparable to that of very young hearing children. Cochlear
implanted children performed significantly better than all the
other groups, and the less accurate performance was that
of foreigner deaf students, who often have not any kind of
underlying language. Despite the better performance of cochlear
implanted children, who generally do not use the sign language,
the best solution to approach the oral language would appear
to be the combination of oral training and sign language, in
order to be able to communicate with both the deaf and the
hearing communities. The school system in this sense should
find some strategies in order to help deaf foreigners to get
access to the grammar of the oral language. 
La sordità è una minorazione sensoriale che incide gravemente
sulla normale acquisizione del linguaggio e sullo sviluppo delle
abilità linguistiche. Le persone sorde, non avendo accesso
diretto all’input linguistico, sono limitate nel loro sviluppo della
lingua orale tanto che molti non acquisiscono che i rudimenti
della comunicazione orale. Allo stesso modo dei bambini udenti,
che acquisiscono spontaneamente e naturalmente la lingua
parlata, i bambini sordi possono acquisire la lingua dei segni,
che si serve del canale visivo.
Questo studio esamina la competenza linguistica
dell’italiano, in quattro differenti gruppi di individui sordi
(bambini con impianto cocleare, segnanti nativi, segnanti non
nativi e sordi stranieri adolescenti e adulti), studiata attraverso
la somministrazione di uno specifico test standardizzato di
misurazione della competenza linguistica, e che si avvale
dell’associazione di immagini a frasi. I risultati provano che
molti soggetti sordi mostrano una prestazione paragonabile a
quella di bambini udenti molto più giovani. I bambini con impianto
cocleare mostrano una prestazione considerevolmente migliore
che negli altri gruppi, mentre la prestazione meno accurata
è stata data dagli studenti sordi stranieri i quali spesso non
possiedono nessuna lingua di base. Nonostante la prestazione
migliore è risultata essere quella dei soggetti impiantati, che
generalmente non usano la lingua dei segni, la migliore soluzione
per un efficace approccio alla lingua orale, sembra essere la
combinazione di apprendimento della lingua orale e uso della
lingua dei segni, al fine di comunicare con la comunità sorda e
la comunità udente. In tal senso il sistema scolastico dovrebbe
trovare le strategie più adeguate per aiutare i sordi stranieri a
sviluppare la grammatica della lingua orale del paese che li ospita.
Key words: deafness, oral training, sign language, Italian,
school teaching, language acquisition, language learning. 
* [email protected]
52
Introduction
Children acquire language spontaneously and
effortlessly. They do this in a surprising way and
they are able to master completely the language to
which they are exposed within a period of few years.
Children have innate language-specific abilities
that allow language acquisition to take place in
the first years of life during which environmental
exposure is fundamental to stimulate this innate
ability [1], [2], [3]. It is therefore necessary for this
innate component to be stimulated within a specific
period, known as ‘critical period’, the end of which
is identified with puberty, or otherwise it becomes
more difficult to acquire a language naturally [2].
Some cases of late exposure to linguistic input have
indeed confirmed the critical period hypothesis, as
is the case of Genie [4], who lived confined in a
small room for almost thirteen years. During her
confinement she received no auditory stimulation
and therefore she could not acquire her language
as an infant. She began to learn her first language
late, at adolescence and even if over a period of
years she improved greatly, her mental grammar
remained quite undeveloped.
Also deafness inevitably affects the normal development of speech and language acquisition, since
it drastically reduces both the quantity and quality
of linguistic input available and accessible to the
deaf person. Thus, this has severe consequences
on cognitive and linguistic development [5], which
in most cases persist even after a long rehabilitation process. Various studies investigating linguistic
competence of deaf people found that, although
these individuals might easily learn the lexicon
of a language, they mainly experience difficulties
with most morphosyntactic properties of the Italian
language ([6], [7], [8], [9], [10], [11], [12], [13], [14]).
The most frequent errors in written and spoken
language, as well as in comprehension and production tasks are omission or substitution of determiners, clitic pronouns, prepositions, incorrect use of
number and gender agreement, incorrect use of
verbal morphology, omission of copulas, omission
and/or substitution of auxiliaries and modal verbs.
Deaf learners show preference for shorter sentences
and are less successful in structures that violate
the noun-verb-noun constituency, like in relative
clauses. Also passive constructions are seldom used
by deaf people.
Information and culture, which are transmitted
very largely through language both in the spoken
and written modality, are in most cases precluded
to deaf people, since the difficulties they experience
do not only depend on sensory deprivation but also
on lack of linguistic competence.
Deaf people might acquire and develop naturally
a sign language in which meaning and linguistic
information is not acoustically conveyed, but conveyed using signs which combine simultaneously
hand shapes, orientations, positions and movements
of hands, arms, body and facial expressions. These
languages exploit the visual modality and for this
reason they represent the most natural languages
of deaf communities. Indeed, when deaf people
interact with each other within their community,
it is natural that they use the sign language as
the primary means of communication [15]. As a
consequence, the only possibility deaf children have
of being exposed to a kind of language is the use
of sign language. Sign languages are spoken by
small groups of individuals. They have the same
characteristics as oral languages, i.e. they have their
own grammar and they vary cross-linguistically (for
instance, we mention the American Sign Language
(ASL), the British Sign Language (BSL), the Estonian Sign Language (ESL), the Indo Pakistani Sign
Language (IPSL)).
On the other hand, deaf individuals are surrounded by hearing people communicating via an
oral language, which they have to learn and use in
order to avoid isolation from the “world” around
them. Hence, both the sign language and the oral
language are essential for the deaf individual in
order to have an effective communication system
with both hearing and deaf populations.
This study aims to explore the acquisition and
development of the Italian language by four different
groups of deaf individuals, in order to identify the
main difficulties experienced by these populations in
interpreting different types of sentences of the Italian
language, by using a standardized comprehension
test. We would like our results to help provide
information to account for the difference, in performance, between the four groups in order to awaken,
as many people as possible, to the problems raised
by deafness in the teaching of oral languages.
Deafness in Italy and across the world
In Italy, approximately one out of 1000 people
is born with a hearing loss [16]. Over 94% of deaf
children have hearing parents and the remainder
are children born to deaf parents [17].
Oral language and sign language
Educationally, deaf people constitute a very heterogeneous group. Those who are born deaf or
whose deafness occurs before the age of two or three
may be described as ‘prelingually deaf’. Deafness
occurring after that period is defined as ‘post-lingual’
deafness. Then, we distinguish those who know and
use Italian Sign Language (henceforth LIS) and those
who do not, those who are trained orally and those
who have approached the language either through
the bimodal method or through bilingual education
(see the next section for a detailed description of
these approaches).
Deaf people born to deaf parents acquire sign
language as their first language (native LIS signers),
whereas oral language might constitute for them the
second language and is usually learnt after a period of
intensive training. They usually do not wear cochlear
implants. Native LIS signers form part of the ‘Deaf
Community’ and are mostly proud of their language
and of their culture. For their children, they claim
their right to have a “language of communication”
as well as a “language of scholarly education”. Only
5-10% of deaf children can learn sign language
naturally from their deaf parents. Generally most
deaf children are born to hearing parents and are
not exposed to sign language from birth. Hence, for
them, it is difficult to determine their first language
(either oral or signed), if they actually have any.
Indeed, they could be considered as having no actual
first language, or only a partial one, depending on
the degree of hearing loss and, eventually, on the
age of first exposure to sign language. There are
deaf children that approach sign language after 3-6
years old (early signers). They learn it from other
deaf children when they begin school, in a special
school or residential schools where deaf students
are introduced in classes with other deaf students,
but also live in a boarding arrangement for a long
period.
There are also many deaf people that approach
sign language after 12 years old (late signers).
Deaf people who are not exposed to a sign language early or until adulthood, never do as well
as those who learned it as children [18]. In 1990,
Newport’s study on the acquisition of language and
of American Sign Language (ASL) in adulthood,
revealed that the ASL of people exposed after 12
years old has more flexible morphological rules if
compared to the ASL of native or earlier speakers.
In the same way, in Italy we can find people
with different levels of linguistic competence of
LIS depending on the age of introduction to this
53
language and on the LIS level of the people with
whom the subject interacts.
A phenomenon that is characterizing Western
Europe is that the society is undergoing a radical
change due to the emigration of poor populations
to rich countries. Schools have to host an increasing
number of foreign children, among whom a high
number are disable individuals. The presence of deaf
foreigners has consequently raised other problems
on the correct way of providing them with linguistic
competence and to get them integrated both in the
hearing and in deaf communities. Unfortunately
no data exist, to our knowledge, on the problems
raised from this phenomenon and on the linguistic
competence of deaf foreigners learning Italian. As
a consequence of their different family background,
from a linguistic, social, economical and cultural
point of view, their linguistic competence, both in
Italian and in the sign language, feels the effect of
all these variable.
Basically, the factors that influence deaf individuals’
language development are numerous and complex;
among them is the age of the onset of deafness and
its detection, the severity of hearing loss (mild, from
26 to 40 dB, moderate from 41 to 70 dB, severe from
71 to 90 dB and profound greater than 90 dB) the
age of the first intervention, the parents’ linguistic
background (whether they are native signers or
not), the parents’ choice on the approach providing
the child with the linguistic input, the degree to
which parents simplify the input to the child and
the quality of parent-child interaction. According
to the educational philosophy of parents, a deaf
child may receive language input consisting of oral
speech, some form of manual coded language or
sign language [19]. Nonetheless, the linguistic input
that is given to deaf children is often poor [20]. In
the case of deaf foreigner, to all these variables that
deeply influence the language development of deaf
individuals, the different socio-economic status and
linguistic background compound their difficulties in
the acquisition of any language. For this reason, in
our study, we also want to investigate the linguistic
competence in Italian of a group of deaf foreigners,
who have been living here for some time, in order
to determine whether their performance differs
from that of deaf children born to Italian parents.
Nevertheless, in our experiment we have observed
that Italian linguistic competence, in Italian, of deaf
foreigners, that have lived in Italy since they were
6, can be comparable to the Italian deaf children
born to hearing parents.
54
Linguistic background and methods for
developing language
The level of linguistic competence a deaf child
manages to reach in their own language is influenced
by many variables, among which the type of input
they receive and the way in which they have access
to it play an important part, also strongly depending
on the parents’ linguistic background. At present,
various language learning methodologies are available to make language accessible to deaf people:
1. the oralist method
2. the use of Sign Language
3. the bimodal method
4. bilingual education
5. logogenia
1. Oralist method. This method employs exclusively
written and spoken language without any use of
signs. It aims at developing acoustic training, by
means of cochlear implants or conventional hearing
aids. Conventional hearing aids are external devices
helping the deaf children to exploit their residual
hearing, and mainly to develop lip-reading, which
forms the basis of communication. The cochlear
implant is instead a device that is surgically implanted
in the inner ear (in the cochlea) and is activated
by an external device, worn outside the ear. Conventional hearing aids and cochlear implants have
different functions. While the former usually amplifies
sounds, the latter stimulates the auditory nerve,
thus allowing deaf individuals to receive sounds. It
is worth pointing out that due to the high cost of
cochlear implants, these devices are mainly used
in rich countries. Parents who choose this kind of
approach basically exclude the teaching of sign
language because they believe that avoidance of
sign language and oral speech presentation would
result in improved spoken language acquisition.
2. Use of sign language. Sign language is a visualgestural language, which is considered a full-fledged
natural language. Linguistic research has demonstrated that it has the same degree of expressiveness
and grammatical complexity as any other language
in the world [21] and the development of grammar
rules in sign language follows the same processes as
acquisition of an oral language by hearing children.
As already explained in the previous paragraph,
sign language represents the first language for deaf
children, mainly for those born to deaf parents.
Sign languages are the most natural languages of
deaf communities and, if we consider that they
are spoken by a small group of individuals, they
are comparable to local languages. Those who use
exclusively sign language tend to reject the oralist
method of teaching language.
Recently, the new professional figure of deaf
educator, with specific competence in teaching sign
language to the deaf children and their family, has
been introduced. In some cases, however, many
families are discouraged from learning the sign
language. Indeed they want to remove the handicap
eliminating all that can make it evident [31].
3. Bimodal approach. It combines the oral and
the visual-gestural modalities, but is fundamentally
based on a unique language (in the case in point,
Italian) [22] [23]. Thus, in interactions, words are
accompanied by signs, maintaining the word order of
the oral language and, for those functional elements
that have not an equivalent sign (i.e. articles, prepositions, plurals, inflected morphemes), deaf people
use some invented signs and the fingerspelling
alphabet. Those who support this approach [22],
[23], [24], [25] claim that the use of the visual-gestural
modality may be useful to improve the acquisition
of a spoken language [26] .
4. Bilingual Education. Bilingualism involves the
knowledge and the regular use of two or more
languages to the same level. In the case of deaf
individuals, it consists in the simultaneous exposure
to both oral and sign language. The main assumption of bilingualism is that there is the possibility
of deaf children acquiring a sign language in the
same way as hearing children acquire an oral one,
therefore this will undoubtedly bring them some
advantages in the developmental process and in
the development of an oral language. Hence, deaf
children will be able to meet their own needs, that
is, to communicate early with the people surrounding them, developing cognitive abilities, acquiring
knowledge of the world and getting acculturated
into the world of the hearing and of the deaf.[27],
[28], [29], [30], [16], [31], [32], [33], [34].
5. Logogenia [9]. This method is strictly written
and exploits the reading ability of the children, to
teach them some properties of the language, making
use of strategies like minimal pairs and commands.
This method substitutes the sentences they cannot
hear with written sentences for them to read. Since
it exploits the reading ability, it can only be adopted
at a later stage in language development, at a point
when this ability is available in the deaf child.
Independently from the intervention approach
adopted, every deaf individual seems to be unique
as far as their level of competence, both in oral and
in sign language. Each of these interventions gives
different results also depending on the person and
the intuition of the speech therapist and/or educator.
Deafness and school education
The outcome of speech therapy varies crossindividually, also depending on the educational
system adopted and on the way language is taught at
school. Indeed a good teaching method might help
the deaf person to develop good linguistic abilities.
Deafness raises important problems as far as the
learning of Italian and the educational system selected
to help to teach oral language to deaf people.
Deaf individuals may be introduced in normal
schools, in which they attend classes with hearing
peers, or in special schools, in which all students
are deaf [16].
The main problem deriving from the introduction
of deaf people into hearing classes concern the need
for them to communicate and to get integrated with
the other (hearing) students, as well as to learn the
subjects taught at school. Unfortunately, in most cases
the Italian school is not adequately equipped to
meet deaf people’s requirements. Indeed educational
tools and programs are mainly conceived for hearing
students rather than for deaf ones, who are often
treated as mentally retarded, risking isolation from
the rest of the class.
In special schools, the situation is far from being
better. Although LIS is largely accepted in this type of
school as means of communication, it is nonetheless
not taught formally, i.e. most teachers do not use it
during classes, because they are not native signers,
nor have they attended courses in order to learn it.
Teachers and school tutors are not adequately trained
to teach to deaf people. They use total educational
systems that rely on lip-reading and only sometimes
they use some signs of LIS, action theatre, reading of
textbooks with images, study of grammar to support
the oral language. Moreover, educational programs
do not differ from those in normal schools, since
school subjects in both cases are taught orally, with
strong consequences on the actual comprehension
of the topics presented during classes.
LIS is used by children and students to interact with
their schoolmates and friends in informal situations.
Sometimes, non-native LIS students use instead a
sort of pidgin language. Pidginization is a process
55
occurring as consequence of “relative access to
the target model, the lack of mutually intelligible
language among interlocutors, an immediate need
for communication, and interruption of access to
one’s native language” [35], that is, a code using the
most common LIS signs and many iconic signs, but
without any specific rules.
The advantage of attending a special school is
that deaf students may easily communicate with
their peers, but the disadvantage is that they may
feel excluded from the hearing community and from
its social rules.
Sometimes, special schools attendance may take
place very late, after the failure of methods devoted
to oral training, again with heavy consequences on
the development of linguistic abilities.
Among the different methods to help deaf people
approach oral language, the most accessible form
providing deaf people with grammatical information
to access the oral language, is represented by the
written modality, however, written language has
some limitations, since it is an artificial system, which
excludes every kind of phonological information [12].
Some system are based on the study of grammar
(e.g.[36]). These involve the mastery of specific
metalinguistic abilities, namely a certain degree
of knowledge of the grammar and of the rules
governing it, and a maturity in the mother tongue
in order to understand the description of language.
In this case, the early use of a sign language would
probably help the deaf person to reflect on the rules
governing grammar, in order to apply them in the
acquisition of the oral language.
Grammar and linguistic information might be conveyed through intensive reading [37] [32].Nonetheless,
reading a text also requires certain abilities, namely to
be able to carry out semantic analysis of the natural
language and to make semantic inference; to have
syntax knowledge in order to be able to build the
relationship between the sentence constituents; to
identify time, places and participants involved in the
action; to understand the figurative sense; to identify
the most important elements and capture the core
idea of the text (for more information in Italian see
[38]). The mastery of these abilities requires a high
degree of linguistic competence.
For the teacher, it is very difficult to set and to
control the ability in every child, and to thoroughly
consider every component that plays a role in the
reading. For deaf children, it is very frustrating to
read a text and not to understand it or to misunderstand it. It is necessary to pay a lot of attention
56
to calibrating the difficulty of the text and how
interesting the text is for the children.
Many teachers also adopt grammar texts and
teaching methods for foreign learners. Nonetheless,
again a mother tongue is needed to transfer the
knowledge from the first language to the second
language. The problem arises mainly because often
deaf people cannot transfer such information due to
the lack of a first language. For this reason, a mere
communication skill, like the use of a sort of pidgin
language, is not sufficient to develop language and
linguistic rules.
Our aim is not to discuss here the functionality
of these methods, which sometimes also produce
excellent results, but we want to point out the
necessity of being aware that all the above mentioned systems are thought in order to offer the
language in an alternative way to the natural one.
The artificiality of these systems has consequences
on the level of linguistic competence achieved by
the deaf individuals.
Experimental study: methodology
Table 1 – Participants in CI group
ID
AGE
PARENTS
USE OF LIS
S10
S11
S12
S13
S14
S15
S16
S17
S18
S19
10;8
7;11
9;0
9;6
9;6
8;10
9;5
9;9
9;3
8;1
HEARING
HEARING
HEARING
HEARING
HEARING
HEARING
HEARING
HEARING
HEARING
HEARING
NO
NO
NO
NO
NO
NO
NO
NO
NO
NO
SPEECH
THERAPY
(in years)
9
7
8
9
9
8
8
9
8
7
In the NATIVE LIS group, all participants were
profoundly deaf since birth, born to deaf parents.
They were native speakers of the Italian Sign Language and were hosted in a residential school for
deaf people, in Padua. Two of them habitually used
conventional hearing aids. The most relevant data
concerning this group are shown in the table (2):
Table 2 – participants in the native LIS signers group
Participants
ID
AGE
Thirty-seven deaf individuals participated in this
investigation. They were differentiated into groups
of deaf children using a cochlear implant (CI, N=10;
age range: 7;11-10;8; mean age: 9;2), a group of
deaf adolescent native signers of the Italian Sign
Language (NATIVE LIS, N=7, age range: 13;-17;6;
mean age: 15;11) ([13] Grosselle, 2008), a group
of non-native signers, earlier or late learners of the
Italian Sign Language (NON-NATIVE LIS, N=10; age
range 15;10-24;6; mean age: 18;10) and a group of
foreign deaf individuals (FOREIGNER, N=10; age
range 13;0-24;10, mean age: 17;0).
In the CI group, all participants had profound
hearing loss. They were deaf since birth and had
hearing parents. They had been trained orally and
they had never been exposed to the Italian Sign
Language. They were fitted with hearing aids within
the second year of life and/or they were fitted with
cochlear implants within the third year of life. At
the time of testing, they were receiving speech
therapy two to three times per week and they
were attending primary schools in hearing classes.
None of the children had associated disabilities.
The table (1) contains a summary of each child’s
most relevant data.
S1
S2
S3
S4
S5
S6
S7
15;9
16;1
16;11
15;5
16;5
17;6
13;7
LIS
COMPETENCE
NATIVE
NATIVE
NATIVE
NATIVE
NATIVE
NATIVE
NATIVE
PARENTS
DEAF
DEAF
DEAF
DEAF
DEAF
DEAF
DEAF
In the NON-NATIVE LIS group, all participants
were deaf since birth, born to hearing parents.
They were exposed to the Italian Sign language
very late and they managed to achieve relatively
good competence in it. Some of them received
speech therapy when they were younger. They
wore hearing aids only at school during classes.
The most relevant data concerning this group are
shown in the table (3):
The group of foreign deaf individuals was a heterogeneous group including people ranging in age
between 13;0 and 24;10 years. They were deaf since
birth, born to hearing parents. They mainly came
from the Eastern Europe and belonged to families
that were not able to provide for them. Most of them
had been living in Italy for at least two years. They
arrived here at different ages, some of them arrived
57
Table 3 – participants in the non-native LIS signers group
ID
AGE
(Y;M)
YEARS OF
LIS USE
LIS
COMPETENCE
S20
S21
S22
S23
S24
S25
S26
S27
S28
S29
15;10
16;2
17;5
18;3
19;0
19;5
19;6
20;6
24;6
18;2
10
7
14
7
12
5
10
16
18
13
VERY GOOD
VERY GOOD
VERY GOOD
VERY GOOD
VERY GOOD
GOOD
GOOD
VERY GOOD
GOOD
VERY GOOD
here when they were very young (about 5 years old)
and others arrived when they were between the age
of 10 and 20. Four of them attended school in hearing
classes. With their hearing peers, they attended practical subjects (musical education, technical drawing,
physical education and drawing), they then followed
differentiated teaching during English, mathematical
and sciences. Six students attended instead a special
school for the deaf. They wore hearing aids only
during classes. They showed difficulties in getting
integrated into classes with hearing students because
of the lack of communication skills both in Italian
and in sign language. Moreover, because of their low
communication skills, it was not possible to obtain
some data concerning their hearing loss and their
experience. The available relevant data concerning
this group are shown in the table (4):
Procedure
All participants were tested individually in one or
more sessions, in a quiet place. Deaf children using a
SPEECH
THERAPY
(IN YEARS)
6
6
1
NO
10
10
14
NO
5
2
PARENTS
HEARING
HEARING
HEARING
HEARING
HEARING
HEARING
HEARING
HEARING
HEARING
HEARING
cochlear implant were tested by the speech therapist
and the second author during their individual speech
therapy sessions, while the individuals included in
all the other groups were tested by the first author
at their school.
For children with cochlear implant, the test was
presented orally, while for the other deaf groups
the written modality was preferred. Sentence stimuli
were presented on separate strips of paper in order
to avoid difficulties due to incorrect lip-reading.
Each participant was presented with some pictures
and after the stimulus was read, the participant
had to point to the correct picture. For implanted
children, the sentence was read by the experimenter,
whereas for the other groups, the stimulus was read
autonomously by the participant. Before beginning
the experiment, the correct comprehension of lexical words was verified in order to make sure that
participants were familiar with the names and verbs
presented in the experimental trials.
Table 4 – participants in the foreign deaf group
ID
AGE
(Y;M)
YEARS IN IT
LIS
COMPETENCE
S30
S31
S32
S33
S34
S35
S36
S37
S38
S39
18;2
13;0
17;2
17;11
24;10
17;8
15;0
16;3
14;8
15;7
2
2
2
2
4
5
5
4
8
10
VERY GOOD
VERY LITTLE
LITTLE
VERY LITTLE
VERY GOOD
LITTLE
VERY GOOD
LITTLE
GOOD
GOOD
SPEECH
THERAPY
(IN YEARS)
3
2-3
1
UNKNOWN
NO
NO
2
NO
3
3
PARENTS
HEARING
HEARING
HEARING
HEARING
HEARING
HEARING
HEARING
HEARING
HEARING
HEARING
58
Materials and score attribution
The test used to assess the linguistic abilities of these
participants is a standardized test known as TCGB
(Test di Comprensione Grammaticale per Bambini
‘Test of Grammatical Comprehension for Children’)
[39]. The test TCGB is standardized on hearing people
and it is used to assess the development of children’s
comprehension abilities from 3;6 to 8 years. Unfortunately, at least in Italy, there are not linguistic tools
elaborated for and standardized on deaf people.
Nonetheless this tool is useful in order to provide a
picture of language evolution in terms of linguistic
age. Through response scores for each sentence
typology, it is possible to identify the processes and
the strategies underlying some aspects of the Italian
grammar and to identify vulnerable linguistic areas.
The test includes 76 sentences and, for each trial,
four pictures were shown to participant.. After the
stimulus is proposed, subjects were invited to point
to the picture that correctly matches the sentence,
out of the four possible choices.
Eight different sentence typologies were investigated: items containing locative complements (e.g.
La palla è tra il tavolo e la sedia ‘the ball is between
the table and the chair’), items testing verbal and
nominal inflectional morphology (e.g. camminano
‘(they) walk’, bambino ‘child.masc’), affirmative
active sentences (e.g. la mamma lava ‘the mum
washes’), negative active sentences (e.g. il bambino
non dorme ‘the child does not sleep’), affirmative
passive sentences (e.g. il cane è morso dal bambino
‘the dog is bitten by the child’), negative passive
sentences (e.g. la mela non è presa dalla bambina
‘the apple is not taken by the child’), relative clauses
(e.g. il babbo tiene il palloncino che il bambino
rompe ‘the dad holds the balloon that the child
breaks’), sentences containing dative complements
(e.g. il babbo porta le sigarette al bambino ‘the dad
brings the cigarettes to the child’)
For each response, an error score is attributed.
Scores were attributed in the following way. Each
correct response was attributed 0 scores. If after
the first administration, the participant failed to
provide the correct response, the sentence was
proposed again. When at the second administration,
the participant pointed to the correct picture, a score
of 0.5 was assigned. When they pointed again to
the incorrect picture, a score of 1.5 was attributed.
The final total score was obtained by summing all
partial scores.
For each of the sentence typologies investigated
as well as for the overall performance, the TCGB
manual provides normative data collected from
typically-developing children.
On the basis of these data, it was possible to
attribute a linguistic age to the participants of this
experiment.
Results
The total scores of each participant in each group
are shown in the table (5), also including the mean
score and the standard deviation for each group.
The comparison of our results with normative
data shows that in the CI group, in most cases,
the performance is comparable to that of children
from 5;6 to age peers. In the NATIVE LIS group,
the overall performance was comparable to that of
Table 5 – Total TCGB scores for each participant in each group
CI GROUP
ID
TCGB
S10
8,5
S11
6
S12
13,5
S13
4,5
S14
0,5
S15
8,5
S16
2
S17
0,5
S18
1,5
S19
3,5
M
4,9
SD
4,24
NATIVE LIS GROUP
ID
TCGB
S1
15,5
S2
4
S3
15,5
S4
16
S5
6,5
S6
11,5
S7
21,5
M
SD
12,9
6,04
NON-NATIVE LIS GROUP
ID
TCGB
S20
32
S21
16,5
S22
40,5
S23
46,5
S24
20,5
S25
24,5
S26
29
S27
27,5
S28
34
S29
28,5
M
30,0
SD
8,91
FOREIGNER GROUP
ID
TCGB
S30
15,5
S31
54,5
S32
58
S33
69
S34
32
S35
66
S36
46
S37
73
S38
34
S39
35,5
M
48,4
SD
18,85
children ranging in age from 5 to 7;6. In the NONNATIVE LIS group, the performance was comparable
to that of children younger than 3;6 till the age of
5;6. In the FOREIGNER group, the performance was
comparable to that of children younger than 3;6 till
the age of 5;5. Although the participants in the CI
group and those in the NATIVE LIS Signers group
show the same linguistic age, it is worth pointing
out that the chronological age of the former group
is much lower than that of the latter group (cf.
Tables nr, 1 and 2.)
Statistical analyses were performed using the SPSS
statistical software package. We ran a between-group
analysis in order to compare the performance of
each of the four groups against the others, by using
the non-parametric Mann-Whitney test for independent samples, since the assumption of normal
distribution of the population was not met in this
case. We carried out various comparisons, trying
all possible combinations between pairs of groups.
The analysis revealed that the CI group is the most
accurate. The CI group performed significantly better
than the NATIVE LIS group (U=9.000 p=0.011), the
NON-NATIVE LIS group (U=.000 p=.000) and the
FOREIGNER group (U=.000 p=.000). The NATIVE
LIS group, which achieved the media total score
of 12.9, performed significantly better than the
NON-NATIVE LIS group (U=2.000 =.001) and the
FOREIGNER group (U=3.000 p=.002). Finally, we
found that that the NON-NATIVE LIS group performed significantly better than the FOREIGNER
group (U=18.500 p=.017). On the basis of these
analyses, it is possible to establish a classification
of the four groups from the one that showed the
most accurate performance to that showing the less
accurate performance: implanted children, native
LIS signers, non-native LIS signers and foreign deaf
students.
Discussion
The present study provided evidence confirming
previous data on the difficulties experienced by
deaf people in the use of functional elements and
in the acquisition of some properties of the Italian
language.
The analysis of responses revealed that the deaf
children included in the CI group mainly follow the
same pattern of performance of typically-developing
children as far as the choice of responses is concerned. Hence, structures that develop at a later
stage in hearing children, i.e. passive sentences,
59
relative clauses, resulted more problematic than other
structures for some children with cochlear implant.
Some interesting remarks are found in the performance of NATIVE LIS signers, NON-NATIVE signers and
in the FOREIGNER group, for whom we identified
nonetheless a quite common pattern of response.
The fact that the test is not standardized on deaf
individuals highlighted that these groups chose some
response strategies that are not observed in hearing
children. The most interesting aspect was that they
mainly relied on linear word order or errors due to
different kind of interferences which will be analyzed
in detail in this paragraph.
The lack of linguistic competence in deaf people
leads them to adopt different strategies to interpret
sentences.
The most important difficulties shown by deaf
people are found in the interpretation of passive
sentences, especially the reversible ones (e.g., la
mamma è presa in braccio dal bambino ‘the mother
is picked up by the child’, or il cane è morso dal
bambino ‘the dog is bitten by the child’, or il cane è
tirato dal bambino ‘the dog is pulled by the man’).
Passive sentences are structures with non-canonical
word order, in which the patient/beneficiary of the
sentence becomes the linear subject, whereas the
agent becomes the indirect object, introduced by the
preposition ‘by’. When both nouns can potentially be
the subject of the sentence, meaning is conveyed not
by semantic plausibility, but by syntactic structures
and functional elements. Reversible passive sentences
proved to be extremely problematic for all groups of
deaf individuals, also including cochlear implanted
children, who nonetheless performed overall significantly better than the other groups, Many deaf
children are trained to comprehend passive sentences
and in most cases they are able to correctly interpret
irreversible passive sentences, in which the linear
subject might be an inanimate noun (e.g. la mela è
mangiata dalla bambina ‘the apple is eaten by the
child.fem’). However, in most items, either passive
or active reversible sentence, the interpretation of
a sentence is mediated by knowledge of a world
labelled by linguistic knowledge and it is not a
spontaneous answer. Hence, for instance, if the
sentence “the child bites the dog” is grammatically
correct, in the knowledge of world, it is less common
than “the dog bites the child”.
Many of them are not able to derive the meaning
of the sentence using functional words. Therefore,
they only consider the lexical words, by omitting
functional words, and adopt specific strategies to
60
interpret the experimental sentences. For example,
to interpret these sentences the deaf groups strongly
rely on the linear word order. The two sentences
containing the preposition tra (between) may help
to explain this phenomenon. Indeed, in the two
items la palla è tra il tavolo e la sedia (the ball is
between the table and the chair) and Il bambino è
tra il babbo e la mamma (the child is between the
father and the mother), most participants pointed
to the image in which the order of the objects was
respectively: ball-table-chair and child-father-mother.
Hence, the order of the objects in the picture reflects
the linear order of the words in the sentence. This
explanation was also provided by the participants,
when at the end of the task they were asked to
give a reason for their choice.
In another experimental trial, the understanding of the prepositions da (from) and a (to) was
investigated, and in this case a different strategy was
adopted. In the sentence l’uccellino vola dalla casa
al nido ‘the bird flies from the house to the nest’, a
participant explained that since the nest is the bird’s
home, the bird was flying towards it. She pointed
to the correct picture, an image in which the bird
moves toward the nest, in this case the presence of
the building (the house) was not important. So the
correct answer is due to a different interpretation
respect to the grammatical sense.
These two examples show that only through lexical
words and semantic plausibility, these participants try
to derive the meaning of the sentence, on the basis
of some knowledge of the world, independently of
the syntactic information conveyed by functional
(semantically empty) words. These meanings are
nonetheless not shared by any other speaker.
As [12] pointed out, the lack of specific functional
elements or inflectional forms in deaf people’s written productions does not means that these people
do not have any mental grammar. Linguistic rules
and grammatical categories are present in the deaf,
and the difference in production, between deaf and
hearing children, could be due to the fact that they
adopt linguistic properties which are not grammatical
in Italian, but are available in other languages.
There are cases in which, the sentence interpretation does not depend on the level of linguistic
competence reached in Italian, but it is due to
interferences of the LIS. A positive interference is
found, for examples, in those sentences investigating verbal inflection, namely those requiring the
identification of an action which is concluded (e.g.
Il bambino ha fatto il bagno ‘the child has had the
bath’). In LIS, the conclusion of an action (perfect
tense) is realized with a specific sign after the verb,
meaning ‘done, finished’, which is co-articulated
with the labial word fatto (done, finished). This
phenomenon is also confirmed by the fact that a
participant belonging to the foreign group, who has
been living in Italy since 2007, and above all having
little knowledge of LIS, did not choose the correct
picture. The picture, that was used to investigate
the perfect past tense, was also used to ascertain
if participants understood the present and future
tenses. The positive LIS interference in the perfect
tense is confirmed by the fact that, thanks to the
word “fatto”, the participants are able to distinguish
the perfect tense from the other tenses and, for
this reason, they provided the correct response. In
the case of perfect tense without the word fatto,
participants responded with incorrect answers to
each tense (present, past and future tense).
Another type of error might be due to the graphical similarity of some Italian words, namely that
between tra (between) and tre (three), which caused
some confusion in the mind of the participants and
hindered them from giving the correct response.
Another case of graphical interference is due to
the homophony and homography between the past
participle of the verb leggere (to read), which is
letto (read.PastParticiple), and the word for ‘bed’
(letto), in the stimulus Il libro è letto dal bambino
‘the book is read by the child’. By trying to find
the correct picture matching these sentences, some
participants asked the experimenter where the bed
was, thus proving that they had not understood
the sentence meaning. Often, the inflected form of
the verb causes some comprehension problems,
because deaf people are not able to attribute the
correct grammatical category.
More generally, when the unknown words were
too many in order to understand the sentence, and
participants were not able to establish some kind of
relationship between the elements of the sentence,
those whose performance was comparable to children younger than 3;6 years, pointed randomly to
the pictures or avoided giving any response.
All these above mentioned strategies suggest that
when deaf children read a text, they try to interpret
it by using every piece of information they have at
their disposal, leaving functional elements out of
the computation.
When teaching to deaf children, it is necessary to
take into consideration all these specific problems
shared by most deaf populations.
Conclusion
This study has demonstrated that deafness is a
considerable obstacle to natural language acquisition
and to the mastering of many properties of the
Italian language. For deaf people, and especially for
those who have been exposed late to a language,
the development of grammar seems to be extremely
problematic both in the oral language and in the
sign language. If the use of cochlear implant seems
to be the best device for children to achieve good
competence in the oral language, in some cases the
mastering of the oral language is not yet comparable to that of hearing peers. For deaf people, it is
evident that language acquisition is a non-natural
process, taking place through intensive training and
hard teaching, which represents an artificial and
non-natural system.
We cannot talk about acquisition of language,
which involves knowledge in the natural process
of absorbing and it is context-dependent. In deaf
people, we can mainly talk about language learning,
which requires a deliberate method of achieving
knowledge involving both the active participation
of the learner and a systematic method of teaching;
if the former is a natural and spontaneous process,
the latter requires strength of will.
The study of oral language often requires a lot of
effort. For this reason, probably the best solution is
the use of a bilingual approach, that is, sign language
guarantees the activation of the core grammar which
is necessary to gain linguistic competence in the
oral language. Indeed, the linguistic competence an
individual has in sign language might be transferred
into oral language, through the language provided
by hearing people. It would be necessary to find
easier methods in order to help deaf individuals to
approach the oral language.
At the moment, we cannot expect deaf children
to achieve a linguistic competence comparable to
that of hearing people. Deaf children usually have
to undergo intense training in oral language and
consequently, they do not manage to live all the
experiences that hearing children have with reading
fairy tales, playing sport, playing games, etc. The
use of the LIS makes it possible for deaf children
to satisfy their communication needs and, at the
same time, to develop linguistic competence in oral
language, through a specific training program.
61
Acknowledgments
We thank all our deaf participants, their families,
their speech therapists, educators and teachers.
References
1.
Chomsky N. Reflections on Language. New York: Pantheon Books; 1975.
2.
Lenneberg EH. Biological Foundations of Language. New York: John
Wiley & Sons; 1967.
3.
Pinker S. The Language Instinct: How the Mind Creates Language. New
York: Harper Collins; 1994.
4.
Curtiss S. Genie: A Psycholinguistic Study of a ‘Modern-Day Wild Life’.
Academic Press, New York; 1977.
5.
Furth HC. Thinking without language: psychological implication of
deafness. Free Press; 1966.
6.
Volterra V, Bates E. Selective impairment of Italian grammatical morphology
in the congenitally deaf: a case study. Cognitive Neuropsychology. 1989;
6:273-308.
7.
Fabbretti D, Volterra V, Pontecorvo C. Written language abilities in Deaf
Italians. Journal of Deaf Studies and Deaf Education. 1998; 3:231-244.
8.
Franchi E. Piena competenza e assenza di competenza linguistica. Una
distinzione messa in evidenza dalla Logogenia. Essere copula e ausiliare
in italiano infantile e in un sordo profondo prelinguale non segnante.
PhD Dissertation, Università di Firenze; 2004.
9.
Radelli B. Nicola vuole le virgole. Bologna: Zanichelli; 1998.
10. Ajello R, Marotta G, Mazzoni L, Nicolai F. Morphosyntactic fragility in the
spoken and written Italian of the deaf. In Fava Elisabetta (ed.), Clinical
Linguistic. 2001; 1:101-117.
11. Volterra V, Capirci O, Caselli MC. What atypical populations can reveal
about language development: the contrast between deafness and
Williams syndrome. Language and Cognitive Processes. 2001; 16:219-239.
12. Chesi C. Il linguaggio verbale non standard dei bambini sordi. Roma
Edizioni Universitarie Romane; 2006.
13. Grosselle S. Valutazione della Competenza Linguistica Generale di Adolescenti Sordi Segnanti. University of Venice: Graduation Thesis; 2008.
14. Volpato F, Adani F (in press). The subject/object relative clause asymmetry
in Italian hearing-impaired children: evidence from a comprehension
task. In V.Moscati (ed), Proceedings of the XXXV Incontro di Grammatica
generativa, Siena 2009.
15. 15. Kegl J, Senghas A, Coppola M. Creation through contact: Sign language
emergence and sign language change in Nicaragua. In M DeGraff (edited
by), Comparative Grammatical Change: The Intersection of Language
Acquisistion, Creole Genesis, and Diachronic Syntax. Cambridge, MA:
The MIT Press. 1999;179-237.
16. Maragna S. La sordità. Educazione, scuola, lavoro e integrazione sociale.
Milano: Hoepli; 2000.
17. Lederberg A. Language Development of Deaf Children with Hearing
Parents, in Encyclopedia of Language & Linguistics. 2006;361-368.
18. Newport EL. Maturational Constraints on Language Learning. Cognitive
Science 1990; 14:11-28.
19. Quigley SP, Paul PV. Language and Deafness. College-Hill Press, San
Diego, CA. 1984.
20. Maxwell MM, Falick TG. Cohesion and quality in deaf and hearing children’s
written English. Sign language 1992; 77:345-372.
21. Klima E, Bellugi U. The Signs of Language. Cambridge: Harvard University
Press; 1979.
22. Beronesi S, Massoni P, Osella T. L’italiano segnato esatto nell’educazione
bimodale del bambino sordo. Torino: Omega; 1991.
62
23. Massoni P, Maragna S. Manuale di logopedia per bambini sordi. Milano:
Franco Angeli; 1997.
24. Biscaro B. Orolessia o lettura labio-facciale integrata. L’educazione dei
sordi. 1991; 92/1:35-41.
25. Cuzzocrea R. La dattilologia fonologica bimanuale nello sviluppo della
competenza linguistica dei bambini sordi. Giornale di pedagogia. 2005,
suppl.3.
26. Rinaldi P, Caselli C. Lexical and Grammatical Abilities in Deaf Italian Preschoolers: The Role of Duration of Formal Language Experience. Journal
of Deaf Studies and Deaf Education. 2009;14(1):63-75.
27. Chiri D. Un esperimento pilota di insegnamento bilingue presso una scuola
materna di Cossato. In Caselli MC, Corazza S (Eds.): LIS – Studi, esperienze
e ricerche sulla Lingua dei Segni in Italia – Atti del 1 Convegno Nazionale
Sulla Lingua dei Segni,Tirrenia (PI): Edizioni Del Cerro. 1995;34-41.
28. Chiri D. Educazione bilingue (lingua dei segni/lingua italiana) nella sperimentazione nella scuola d’infanzia “Cossato Centro”. In C Bagnara, G
Chiappini, MP Conte, M Ott, (Edited by) Viaggio nella città invisibile. Atti
del secondo convegno nazionale sulla Lingua Italiana dei Segni. Tirrenia
(PI): Del Cerro. 2000;243-244.
29. Terracchini E, Bocchini V, Pavarotti M, Antonioni S. Esperienza di integrazione di due alunni sordi in una classe di udenti attraverso attività dei
bilinguismo e progettazione ipertestuale. In C. Bagnara, G Chiappini,
MP Conte, M Ott, (eds.) Viaggio nella città invisibile. Atti del secondo
convegno nazionale sulla Lingua Italiana dei Segni, Tirrenia (PI): Del
Cerrro. 2000; 280-286
30. Sansoni M, Gaggiotti M. Lingua dei segni nella scuola media inferiore:
racconto di un’esperienza. In C. Bagnara, G. Chiappini, M.P. Conte, M. Ott,
(eds.) Viaggio nella città invisibile. Atti del secondo convegno nazionale
sulla Lingua Italiana dei Segni. Tirrenia (PI): Del Cerrro.2000; 272-279
31. Caselli MC, Maragna S, Pagliari Rampelli L, Volterra V. Linguaggio e Sordità.
Firenze: La Nuova Italia; 1994.
32. Ardito B, Mignosi E. Vivo una favola e Imparo le lingue. Giocare e parlare
con i bambini sordi e non. Firenze: La Nuova Italia; 1995.
33. Teruggi LA (ed). Una scuola, due lingue: l’esperienza di bilinguismo della
scuola dell’infanzia ed elementare di Cossato. MI:Franco Angeli; 2003.
34. Pinto MA, Volterra V (eds). Bilinguismo lingue dei segni / lingue vocali:
aspetti educativi e psicolinguistici. Sign languages/spoken languages
bilingualism: educational and psycholinguistic issues. Rivista di Psicolinguistica Applicata Special Issue. Pisa-Roma: Fabrizio Serra ed. 2008-VIII/3.
35. Lucas C., Valli C. ASL, English, and contact Signing. In Ceil L, Sign Language
Research: theoretical issues. Gallaudet University Press. 1990; 288-307:302
36. Giuranna R. Analisi grammaticale visuale della lingua italiana. In Caselli
MC, Corazza S (eds): LIS – Studi, esperienze e ricerche sulla Lingua dei
Segni in Italia – Atti del 1 Convegno Nazionale Sulla Lingua dei Segni.
Tirrenia (PI): Edizioni Del Cerro.1995; 34-41.
37. Favia ML, Maragna S. Una scuola oltre le parole. Manuale per l’istruzione
dei sordi. Firenze. La Nuova Italia; 1995.
38. Cornoldi C, De Beni R, Gruppo MT. Guida alla comprensione del testo.
Bergamo, Ed. Scolastiche Walk Over;1989.
39. Chilosi AM, Cipriani P, Giorgi A, Fazzi B, Pfanner L. TCGB. Test di comprensione grammaticale per bambini. Pisa: Edizioni del Cerro; 2006.
LÁBIOS, LEITE, CHOCOLATE, LARANJA, ETC.:
um estudo sobre os nomes das cores em LSB*
LIPS, MILK, CHOCOLATE, ORANGE, ETC.:
a study of colors’ names in Brazilian Sign Language
Sandra Patrícia de Faria do Nascimento1**
1
Universidade de Brasília – LIP
Resumo
Abstract
Os nomes das cores, em todas as línguas do mundo, abrigam
questões lingüísticas muito interessantes e distintas. Por isso,
decidiu-se analisar os processos de denominação das cores
em Língua de Sinais Brasileira (LSB). Gerou-se um corpus a
partir de nomes de cores lexicografadas em dois repertórios
com LSB e a esse corpus aplicou-se o Modelo Sílex. Esses
procedimentos permitiram: (a) identificar, analisar e sistematizar
os processos de construção dos nomes de cores em LSB e,
ainda, (b) confirmar a existência da competência construcional
dos falantes de LSB, como L1, ao preencherem lacunas lexicais
na LSB; (c) constatar que o Modelo Silex sustenta uma análise
morfológica que contempla questões funcionais da língua; e, por
fim, (d) verificar a importância e validade de análises diacrônicas, na
identificação dos processos de construção de palavras em LSB.
Palavras-Chave: Cores. Língua de Sinais Brasileira. Modelo
Silex. Morfologia. 
Introdução
Os nomes das cores, nas mais diferentes línguas,
abrigam questões que vão desde a utilização de um
único nome para várias cores e tonalidades até a
denominação bem discriminada e detalhada das
diversas tonalidades de uma mesma cor. Para analisar
os processos de construção morfológica que levam
a essas questões, há modelos teóricos estruturalistas
excelentes. Para analisar a denominação das cores
em Língua de Sinais Brasileira (LSB), optou-se por
aplicar à análise do corpus gerado para esse estudo,
a segunda versão do Modelo Silex proposto por
Corbin (1997b), que se trata de um modelo mais
funcionalista.
The colors’ names in all languages of the world enclose
very interesting linguistic aspects. We decided to analyze the
processes of colors denomination in Brazilian Sign Language
(LSB) to understand how these processes occur in this language.
The colors’ names corpus was generated from two repertoires
with LSB. Then, the Silex Model was applied to this corpus. These
procedures allowed us: (a) to identify, analyze and systemize
the construction processes of the colors’ names in LSB; (b)
to confirm the existence of the constructional competence of
first language speakers of Brazilian Sign Language, when filling
lexical blanks in LSB; (c) to evidence that Silex Model supports
morphological analysis that contemplates functional aspects
of the language; and, at last, (d) to verify the importance and
validity of diachronical analyses, in the identification of the words’
construction processes in LSB.
Keywords: Colors. Brazilian Sign Language. Silex Model.
Morphology. 
A grande contribuição desse modelo, criado com o
intuito de construir uma teoria própria para o léxico,
é a de conceber uma MORFOLOGIA CONSTRUCIONAL ASSOCIATIVA, na qual a Semântica Lexical e
a Morfologia se unem para descrever fenômenos
formais regulares que levam, em primeira mão, a
uma descrição autêntica do léxico.
Esse modelo elege palavras mais transparentes na
língua e as analisa, indutivamente, de forma a tornar
mais evidente a relação que existe entre a unidade
lexical e a sua referência. Nesse percurso, descreve
como as palavras e seus respectivos significados
se constroem numa língua. Para Correia (1999:3),
o estudo das palavras construídas, por meio desse
* Trabalho desenvolvido durante e após Doutoramento no Departamento de Lingüística, Português e Línguas Clássicas – LIP, da
Universidade de Brasília – Brasil.
** Email: [email protected]
64
modelo, constitui um observatório privilegiado para
a construção do significado.
Outra vantagem do modelo relaciona-se a aspectos extralingüísticos presentes no ato da criação
lingüística, isto é, no ato da denominação lexical.
Em face do caráter icônico1 das línguas de sinais,
o extralingüístico parece mais transparente que nas
línguas orais, apesar de uma análise acurada do
processo de denominação na língua de sinais permitir
a identificação do caráter lingüístico marcadamente
cognitivo em todas as etapas da criação lexical,
quer pela evidência de construções metonímicas
representadas, preponderantemente, pela figura da
sinédoque; quer pela identificação de protótipos na
composição de denominações dos referentes; quer
pelo ícone lingüístico convencionado pelos falantes.
Ao se prestar à descrição das línguas de sinais,
esse estudo das palavras construídas, em LSB, pode
ser extremamente útil à ampliação do léxico tanto
da língua comum como do léxico das diversas áreas
de especialidade. Assim, a aplicação desse modelo
parece possibilitar ao pesquisador, elencar, também,
estruturas contidas no lexicón2 da LSB.
Uma das importantes considerações trazidas pelo
modelo deve ser explicitada para se entender como
o sentido de um referente motiva a construção
das denominações e, ao mesmo tempo, como as
denominações absorvem os sentidos dos referentes
que denominam. O Modelo Silex concebe o sentido
de uma palavra como inerente a ela; é individual e
particular; não é um sentido que resulta no texto e
no contexto. Essa ênfase é necessária para não se
confundir “construção de sentido de uma palavra”
com “construção de sentido de textos”. São dois
processos de natureza diferente. O sentido inicial
de uma palavra emerge dos processos construcionais lingüísticos que podem ser partilhados com
elementos extralingüísticos empregados para a sua
concepção; enquanto o sentido de um texto prescinde do contexto e, nele, as palavras, antes com
um significado prévio, ressignificam.
1
2
É necessário ter bastante clareza quando se fala em iconicidade
nos estudos das línguas de sinais, pois o ícone pode ser visto
em duas perspectivas: como forma e como cognição. Ainda
que a forma esteja transparente na concepção de um item
lexical, em LSB, essa motivação inicial é, por excelência,
cognitiva. Essa asserção é irrefutável, dado que a cognição
está presente em todo o processo de construção de uma
palavra em Língua de Sinais.
O termo lexicón, empregado nesse estudo, refere-se ao fundo
lexical de uma língua, à base de constituintes e de estruturas
lingüísticas internamente organizadas e disponíveis para a
construção lexical de uma língua.
Assim, o Modelo Silex, que tem como escopo
a PALAVRA, advoga que ela tem, sim, um significado lexical inerente, entendido como pré-requisito
partilhado3. O significado lexical, baseado nesse
modelo, não é de todo adquirido no contexto e no
uso. Por isso, nessa concepção, “o significado das
palavras construídas não é alvo de memorização: a
gramática da língua fornece os meios de o calcular”
(CORREIA4, 1999). Nessa ordem, para a compreensão
do significado, então, participam três elementos:
língua (relacionada ao sentido inerente), cultura
(relacionada ao sentido atribuído pelo estereótipo)
e percepção (relacionada à concepção prototípica).
Para alcançar esse mérito, o modelo postula,
então, que há três “significados” intervenientes na
construção do “significado”, por exemplo, de uma
palavra derivada: o significado conferido pela regra
de formação de palavras, o significado herdado da
base; o significado específico do operador morfológico
(CORREIA, 1999).
Processos de construção de palavras –
Modelo Silex
Corbin (1997b) e Correia (1999:65-77) consideram,
para o Modelo Silex, dois grandes processos construcionais: um de produtividade e outro de criatividade.
Neles, são reconhecidas as seguintes operações
morfológicas para a construção de palavras:
(a) como processo de produtividade5:
(i) a derivação, que ocorre por meio dos processos
de afixação e, mais comumente, por meio da
sufixação. Corbin (1997b) e, por conseguinte,
Correia (1999) defendem que o afixo é portador
de uma instrução semântica específica, e a aplicação dele a uma base obedece a determinadas
restrições;
(ii) a conversão (CORREIA, 1999:71-3), que equivale
ao processo tradicionalmente conhecido por
DERIVAÇÃO IMPRÓPRIA, REGRESSIVA, que
3
4
5
A afirmação de que as palavras têm um significado lexical
inerente se contrapõe a estudos propostos pela Lingüística
textual (cf. KOCH, 2000; MARCUSCHI, 2001 e 2002 e outros)
e pela Pragmática (cf. LEVINSON, 1983; MEY, 1993/2001 e
outros). Esses estudos focalizam outra dimensão: a construção
do sentido de textos. Consideram, portanto, que o sentido
não se restringe ao texto e nem é inerente a ele: é construído
na interação autor-leitor.
Correia (1999) é defensora do modelo proposto por Corbin
(1997b).
Produtividade, entendida sob o conceito de Lyons (1977)
como uma característica inerente ao sistema lingüístico.
Lábios, leite, chocolate, laranja etc.: um estudo sobre os nomes das cores em LSB
trata da construção de nomes deverbais a partir
do radical do verbo, por mera adição de um
morfema de gênero. Ocorre em maior proporção
na passagem de ADJETIVOS > SUBSTANTIVOS
e em menor proporção na passagem de SUBSTANTIVOS > ADJETIVOS;
(iii) a composição, que, para CORREIA (1999:701), é a combinação de duas unidades lexicais
ou infralexicais6 de significado descritivo. Ela
entende que o elemento da esquerda se comporta como elemento de composição quando
é um nome, adjetivo, verbo ou advérbio, com
significado descritivo. E se comporta como
prefixo quando o significado é instrucional.
Esse processo altera o significado descritivo da
base, sem alterar o significado referencial. Um
exemplo em português do Brasil, lembrado por
Correia (1999:77), é “beijoca”.
Essa descrição permite a organização do esquema
ilustrativo, mostrado a seguir, com a síntese dos
processos de construção de palavras empregados
no Modelo Silex:
Esquema 1 – Esquema ilustrativo dos processos de construção
de palavras Modelo Silex.
MODELO SILEX
PROCESSOS DE CONSTRUÇÃO DE PALAVRAS
PROCESSOS DE PRODUTIVIDADE
(b) como processo de criatividade :
os processos deformacionais, entre os quais
destacam-se:
(i) a APÓCOPE que consiste na supressão de uma
parte da palavra, o que a torna mais familiar
(CORREIA, 1999:77);
(ii) a AFÉRESE que consiste na supressão de um
segmento no início da palavra; a AMÁLGAMA,
processo pelo qual se constrói uma unidade
lexical pela aglutinação de partes de outras unidades. Como exemplos em português europeu:
borbotixa, nomes próprios antroponímicos como
“Camané” para Carlos Manuel ou em português
brasileiro, “Cantõe” para Carlos Antônio; a REDUPLICAÇÃO DE APÓCOPES e/ou AFÉRESES, que
surge, freqüentemente, associada a onomatopéias.
Também é comum em diminutivos de nomes
próprios como: Juju e Cacá;
(iii) a SUFIXAÇÃO AVALIATIVA APRECIATIVA que
consiste na avaliação qualitativa ou quantitativa
do referente por meio de diminutivos, aumentativos, pejorativos, laudativos etc., expressa
pelo enunciador;
(iv) a SUFIXAÇÃO AVALIATIVA ENUNCIATIVA,
normalmente empregada por um enunciador
próximo ou familiar do referente. Também
é entendida como processo deformacional a
SUFIXAÇÃO FAMILIAR, um tipo de sufixação
avaliativa enunciativa, sempre associada à construção de significados mais enunciativos do que
referenciais, sobretudo em contexto familiar.
7
DERIVAÇÃO POR AFIXAÇÃO
CONVERSÃO
COMPOSIÇÃO
7
Unidades infralexicais são, tipicamente, afixos que não são
sintaticamente autônomos e nem são palavras.
Criatividade, entendida sob o conceito de Lyons (1977) como
capacidade do falante de alargar o sistema lingüístico por
meio de princípios de abstração e comparação imprevisíveis.
PROCESSOS DE CRIATIVIDADE
PROCESSOS DEFORMACIONAIS
AFÉRESE
APÓCOPE
REDUPLICAÇÃO DE
APÓCOPES E/OU AFÉRESES
AMÁLGAMA
SUFIXAÇÃO AVALIATIVA
SUFIXAÇÃO AVALIATIVA APRECIATIVA
SUFIXAÇÃO AVALIATIVA ENUNCIATIVA
SUFIXAÇÃO FAMILIAR
A construção dos nomes de cores em LSB
Como prelúdio à aplicação do Modelo Silex à LSB,
esse trabalho desenvolve, sucintamente, pequenas
reflexões a respeito dos nomes das cores em LSB.
A proposta foi tentar aplicar aos indivíduos lexicais8
selecionados, alguns princípios do modelo, entre os
quais o princípio da teoria da otimidade9, no qual
se acredita que diante da existência de duas regras
em concorrência, a língua não vai adotar a mais
adequada, mas a que melhor funciona no sistema.
Essa análise é bastante interessante, pois justifica
a inexistência de um único paradigma para explicar
os diferentes processos de construção envolvidos
na denominação de indivíduos lexicais de mesma
categoria. A teoria da otimidade dá visibilidade ao
mundo extralingüístico presente na sistematização
dos processos encontrados na análise de denominações das cores, em Língua de Sinais Brasileira, pelo
Modelo Silex, como se vê na aplicação do modelo
na seção que se segue.
8
6
65
9
“Indivíduo lexical” é o nome dado a uma unidade lexical no
Modelo Silex.
A teoria da otimidade tem relação com a proposta gerativista.
Ela dialoga com o Modelo Sílex uma vez que esse modelo,
antes de tornar-se mais funcionalista, estruturou-se sobre
bases gerativas e foi perdendo-as à medida que foi evoluindo.
Manteve-se, contudo, essa proposta da teoria da otimidade.
66
Aplicação do modelo
Por questões práticas e didáticas, a seleção do
corpus deteve-se aos nomes de cores registrados
por Capovilla (2001). Esse repertório apresenta
figuras que visam facilitar a leitura da articulação
das palavras. Foram localizados, na obra, vinte e
seis (26) nomes de cor, entre os quais se encontram
aqueles relacionados a nomes de cores propriamente
ditos e a nomes para a categoria COR(ES) (Figuras
8, 9 e 10 do anexo). Entre esses nomes, estão
relacionadas, ainda, algumas variantes denominativas
como BRANCO (Figuras 5 e 6 do anexo), BEGE
(Figuras 3 e 4 do anexo), VERDE (Figuras 18 e 19 do
anexo), MARROM (Figuras 12 e 13 do anexo), além
de hipônimos do nome verde, presentes em verde
claro e verde escuro (Figuras 20 e 21 do anexo).
O modelo SILEX possibilita, acerca dos dados
selecionados, que se identifique a presença de vários
fenômenos, de diferentes naturezas, para a denominação das cores em LSB. Essa análise encontra
respaldo em Correia (1999:5-6), quando afirma que:
“A denominação caracteriza-se por uma relação
em grande parte não-explicável lingüisticamente.
(...) O sistema lingüístico pode fornecer diversas
denominações para uma mesma classe de entidades (...) A razão que preside a escolha de uma
ou de outra das possibilidades da língua não é,
fundamentalmente, de ordem lingüística. (...) O
fato de escolher uma denominação possível em
vez de outra é demonstrativo do modo como a
categoria de entidades denominadas se dá a ver
no âmbito de determinada língua (CORREIA,
1999:5-6)”.
Para iniciar a empreitada de análise, como um
procedimento viável em um laboratório de Neologia,
buscou-se, primeiramente, identificar se os nomes
de cores, em LSB, ou alguns deles, são fruto de
empréstimos lingüísticos. Esse estudo não se ateve à
identificação de empréstimos por meio da comparação
entre línguas de sinais, pois é um estudo incipiente
e carece de um corpus de várias línguas de sinais
para subsidiá-lo. Esta será, portanto, uma investigação
necessária em futuros estudos a fim de se ter a
certeza de há denominações de cores emprestadas
de outras LS. As hipóteses de empréstimos, nesse
estudo, entretanto, foram levantadas, por meio de
contraste do corpus da LSB com dados de línguas
orais, mais especificamente, da língua portuguesa,
língua em contato mais direto e freqüente com a LSB.
Correia (1999) emprega para o conceito de empréstimo lingüístico o termo importação10. Essa é a
terminologia que também será adotada neste estudo.
Há, basicamente, duas formas bastante produtivas
de “importações” de línguas orais para línguas de
sinais: uma ocorre por meio da “digitação11”, ou seja,
pelo empréstimo das letras equivalentes ao nome
“importado” transliteradas para configurações de mão
correspondentes às letras do alfabeto dactilológico;
a outra ocorre por meio do empréstimo, apenas, da
inicial da palavra da língua oral, totalmente adaptada
e apropriada pela língua de sinais receptora, para
a articulação do item lexical.
Nos casos em que foi constatada “importação”, a
formulação encontrada foi sistematizada. Aos casos
contrários, foram aplicadas as regras de construção de
palavras (RCP) para verificar se as denominações de
cores correspondiam ou não a palavras construídas.
A partir daí buscou-se a identificação dos processos
envolvidos nessa(s) construção(ões).
No rol dos vinte e seis indivíduos lexicais gerados
pelo corpus, composto de designações de nomes de
cor e nomes da própria categoria “cor”, foi identificado uma única denominação cuja forma equivale
à digitação de toda a seqüência de letras do nome
em língua portuguesa; esse foi o caso de C-O-R12
(Figura 8 do anexo). Normalmente, a digitação de
uma palavra da língua portuguesa, pelo alfabeto
dactilológico, significa, além da referência a nomes
próprios, a “importação” de uma palavra, na ausência
de equivalente na LSB ou no desconhecimento da
denominação já preconizada pela comunidade surda
em outra região do país.
Em grande parte das vezes, essa digitação é
temporária; ocorre até que um neologismo ocupe
esse espaço, na LSB. Muitas vezes, palavras curtas
não concorrem com um neologismo na LSB, mas
passam por uma adaptação da estrutura “importada”
à estrutura lexical da língua. Essa adaptação pode
ser manifestada por meio da aceleração rítmica da
10
11
12
“Importação” é o termo preferido por Correia, para as palavras
“importadas” de outras línguas. A preferência de Correia é
altamente justificável, dado o fato de tais palavras não serem
devolvidas à língua de origem, mas, rapidamente, adaptadas
e incorporadas ao sistema lingüístico da língua receptora.
Designamos por “digitação” ao que normalmente é designado
por “soletração” ou dactilologia.
A separação das letras da palavra “cor”, por meio de hífens,
segue o sistema de transcrição de palavras, da LSB, por meio
de notações léxicas. Letras da língua portuguesa, unidas por
hífen, significam que a articulação do indivíduo lexical se dá
por meio da reprodução das letras transliteradas do alfabeto
datilológico, segmentadas por hífen.
digitação da palavra, tornando-a um item lexical com
as mesmas características dos demais itens lexicais
da LSB (cf. FARIA, 2009). Não foi o caso do nome
“cor” que concorre com variantes não importadas
da língua oral, caso de COR II e CORES (Figuras 9
e 10 do anexo).
O fenômeno é idêntico ao que ocorre entre as
línguas orais. Há casos, na LSB, em que a forma
“importada” concorre com o neologismo cujo valor
lexical e semântico é o mesmo. No caso de palavras
concorrentes na LSB, em que uma é “importada”
e outra é construída diretamente em LSB, há uma
grande distinção morfológica entre uma e outra,
pois a forma “importada” foi motivada por letras do
alfabeto da língua oral, transliteradas para configurações de mão da LS e a outra se constitui de uma
estrutura típica da LSB. É o caso de COR I (Figura
8 do anexo) e COR II (Figura 9 do anexo). Parece
haver uma preferência da comunidade surda pela
denominação construída diretamente na língua, sem
motivação na língua oral.
Nos casos em que a palavra é integralmente
“dactilologizada” costuma ocorrer uma acomodação
lingüística na LSB, por meio da aceleração rítmica,
que resulta na lexicalização do termo importado. O
aumento de velocidade na articulação das letras se
acomoda à estrutura de um item lexical em LSB. O
resultado desse processo é conhecido por “soletração
rítmica”. Ocorre uma espécie de assimilação da
palavra “importada” à língua receptora.
Muitas vezes, esse processo gera apócopes, ou
seja, uma palavra digitada e que não concorreu
com equivalente construído na língua pode passar
pela assimilação e, posteriormente, pelo processo
de apócope. É o caso da designação da cor “azul”
(Figura 2 do anexo). A digitação da palavra foi,
inicialmente, “importada” da língua portuguesa, pelo
processo de digitação. Essa denominação sofreu
adaptação morfológica e, depois, apócope das letras
internas Z e U. Era A-Z-U-L13 e passou a A-L (Figura
2 do anexo). Permaneceram a primeira e a última
letra, respectivamente, da denominação em língua
portuguesa.
No corpus analisado, foi recorrente o processo de
construção dos nomes de cores em decorrência de
dois recursos. O primeiro, motivado pela “importação” da letra inicial da palavra equivalente em
13
A representação de palavras, na LSB, por letras separadas
por hífen significa que o sinal (significante) em LSB
está sendo representado pela digitação das letras da
palavra em língua portuguesa.
67
língua portuguesa. Essa letra será a CM utilizada
para a articulação do item designado. O segundo,
motivado pela concepção que o falante tem da
entidade a ser denominada.
Foram identificados no corpus selecionado, onze
“nomes de cores construídos com inicialização”:
cinza (Figuras 7 do anexo), bege I (Figura 3 do
anexo), e bege II (Figura 4 do anexo), lilás (Figura
11 do anexo), marrom I (Figuras 12 do anexo), preto
(Figuras 15 do anexo), roxo (Figuras 17 do anexo),
verde I (Figura 18 do anexo), violeta (Figura 23 do
anexo), castanho (Figura 29 do anexo) e púrpura/
purpúreo (Figura 35 do anexo).
Bege I e bege II representam uma denominação
variante da outra. O mesmo nome designa as cores
‘roxo’, ‘violeta’ e ‘púrpura’, o que torna indistinta uma
tonalidade da outra. Na verdade, em português, a
relação entre roxo e violeta é a mesma, segundo o
Dicionário Aurélio Eletrônico da Língua Portuguesa
(2004). Entretanto, o mesmo dicionário traz púrpura
com equivalência ao vermelho e não ao roxo. Se
‘roxo’ e ‘violeta’ forem analisados em oposição a
‘púrpura’, é possível falar em uma relação polissêmica
entre ambas denominações.
As denominações relacionadas têm, cada uma, a
CM idêntica à da letra do alfabeto dactilológico que
corresponde à primeira letra do nome em língua
portuguesa. Dessa forma, as CMs dos nomes dessas
cores correspondem às letras: “C” para cinza e para
castanho, “B” para bege I e bege II, “L” para lilás,
“M” para marrom I, “P” para preto, “V” para verde
I e “R” para roxo, violeta e púrpura.
À exceção de bege II e castanho, os demais nomes
são, todos, articulados sobre o dorso da mão passiva.
Algumas questões se põem a partir dessa constatação:
o dorso da mão passiva pode ser considerado base
para formação de palavras? A CM que se aloja
sobre essa “base” pode ser considerada um afixo?
Aparentemente, sim. A base é o elemento comum
entre essas formações e o afixo é o elemento que
vai diferir uma cor da outra, que seria o equivalente
a cada CM correspondente à letra inicial do nome
em língua portuguesa. Trata-se, portanto, de um
processo de construção derivacional. Apesar de,
nesse contexto, as CMs “B”, “C”, “L”, “M”, “R” “V”
carregarem o significado dos nomes das cores na
língua da qual as letras foram importadas como
motivação para a construção dos nomes de cores em
LSB, elas, isoladamente, não podem ser entendidas
como elementos com significado. Por isso, não é
possível falar em composição. Assim, essas CMs,
no contexto analisado, têm estatuto de morfema,
68
pois carregam em si um significado suficiente para
serem analisadas como afixos. Posta essa análise,
outra dificuldade que se impõe é a de determinar
se esse afixo diz respeito a um prefixo, um infixo
ou um sufixo, diante do fato de a modalidade da
LSB favorecer a superposição de morfemas. Parece
ser simultânea a articulação da “base” (dorso da
mão) e do Morfema “letra-inicial-importada”. Grosso
modo, por isso, esses morfemas serão denominados de infixos14. Essa análise se confirma se
comparados os indivíduos lexicais registrados por
Capovilla aos indivíduos lexicais registrados por
Oates (Figura 25 do anexo). Resta, então, desses
indivíduos lexicais mencionados como “nomes de
cores construídos com inicialização”, tratar do nome
‘bege II’ e ‘castanho’ que não seguiram o paradigma
de construção por meio da sobreposição da CM
da letra inicial sobre a base. Bege II (Figura 4 do
anexo) e castanho (Figura 29 do anexo) tiveram a
motivação da letra inicial do nome da cor, em língua
portuguesa, mas ambos são articulados no espaço
neutro, com movimento idêntico (movimento lateral
direita-esquerda-direita). Na ausência de um nome
lexicalizado em LSB, falantes de LSB, eventualmente,
usam a letra inicial do nome que aparece em dado
discurso, em dado contexto, como anafórico, nessa
mesma altura, no espaço neutro onde se costuma
digitar qualquer palavra, com o mesmo movimento
ou circular. Percebe-se, ainda, que esse espaço
neutro de articulação das duas denominações não é
arbitrário, pois caso o fosse, poderia cada uma das
cores ser articulada num ponto diferente do espaço
neutro ou do corpo. Trata-se, portanto, de um ponto
de articulação possível, determinado pelos falantes,
fato que tem uma razão de ser, apesar de não ser
possível uma explicação transparente. Também é
salutar lembrar que se ‘castanho’, fosse iniciado pela
CM da letra “C” no ponto de articulação da base
das demais cores, seria um homônimo da denominação para a cor cinza, cor por sinal, em posição
prototípica mais periférica que ‘castanho’. Para evitar
a homonímia, o falante pode ter empregado como
estratégia a locação da denominação em outro ponto
com o fim de eliminar o risco de ambigüidade ao
ser mencionada a cor em LSB.
As categorias de nível superordenado são denominadas, normalmente, em LSB, grosso modo, por um
vocábulo especificador da categoria (cf. FARIA,
2009), normalmente constituído de um elemento
14
Para determinar se são, realmente, infixos ou outro,
cabem muitas análises nesse tópico.
prototípico do nível básico da categoria (ou de
atributo desse elemento), acompanhado do indicador de categoria representado pela glosa “ETC.”
(equivalente às representações das Figuras 27 e 28
do anexo). Há várias categorias que não seguem esse
paradigma. São categorias, aparentemente, opacas.
Numa análise sincrônica e superficial da denominação da categoria ‘cor’, em LSB, identifica-se indivíduos lexicais, aparentemente, diferentes do previsto
no paradigma (Figuras 9 e 10 do anexo). Para uma
análise diacrônica, buscaram-se os repertórios de
Gama (1875) e Oates (1983:165-7), com o intuito de
verificar se eles introduziam a denominação “cor(es)”
em seu repertório. O primeiro não contemplou a
categoria cores. Gama (1875), praticamente, não
nomeia categorias em nível superordenado. Oates
(1983:165-7) inseriu a designação para a categoria
“cores” em seu repertório (Figura 24 do anexo).
Os dados encontrados em Oates facilitaram uma
comparação com o repertório de Capovilla (2001)
e tentou-se, a partir deles, reconstruir um percurso
diacrônico da denominação da categoria ‘cor’, identificada no corpus por cor II (Figura 9 do anexo) e
cores (Figura 10 do anexo).
Foram levantadas hipóteses com base na grande
probabilidade que a LSB tem para denominar
categorias compostas com elementos protótipos
associados a um denominador de categorias, caso
de ETC. (Figuras 27 e 28 do anexo). Foi localizada,
no repertório de Oates (1983), como designação
para a referida categoria, a composição VERMELHO
(Figura 24 do anexo) e o indivíduo lexical que tem
como glosa VÁRIOS, COISAS, DIVERSOS (Figuras
24, 27 e 28 do anexo), mas que pode ser traduzida,
também, por ETC.
Essa estrutura encontrada em Oates (1983) segue
o paradigma da categorização em LSB. Por que,
então, os falantes de LSB teriam rompido com esse
paradigma e construído um novo item lexical para
denominar a categoria, se a lacuna da categoria já
estava preenchida com um termo cuja estrutura é
perfeitamente aceitável pelos falantes de LSB?
A análise diacrônica permite perceber que as denominações encontradas para ‘cor’ não são arbitrárias
e nem romperam o paradigma previsto; elas são
fruto de evolução resgatada na análise diacrônica.
Segue, então, o percurso que este estudo traça, como
hipótese de resgate da construção da denominação
da categoria “cor”.
Os processos assimilativos que atuam nas línguas
deram origem à “teoria do menor esforço”. Segundo
Fromklin & Rodman (1993:335) nessa teoria, “as
mudanças [...] devem-se em primeiro lugar, à “preguiça” lingüística uma vez que ao falarmos fazemos o
mínimo esforço. Por outro lado, os autores ressaltam
também que, embora se verifique simplificação
nas gramáticas, encontra-se também elaboração ou
complexificação.
Entendidos esses processos (de assimilação e
de complexificação) como naturais nas línguas e
aceita a adaptação lexical como forma de sobrepor
informações semânticas sobre um único item lexical,
será possível entender que a tendência dos falantes
de LSB é a de assimilar lexias complexas, tornandoas itens lexicais únicos. É fato que o fenômeno de
complexificação também ocorre em LSB. Para o
momento, entretanto, no corpus de estudo, foram
identificados especialmente os casos de assimilação.
Assim, por meio da alteração rítmica, da alteração
da velocidade de articulação do item lexical15, é
possível supor que a denominação da categoria cor,
presente hoje, na LSB, nada mais é do que o fruto
da evolução dos indivíduos lexicais presentes no
paradigma VERMELHO ETC.
Numa hipótese de reconstituição diacrônica desse
item, é possível dizer que o fato de a denominação
VERMELHO ser articulada próxima aos lábios, atraiu
a articulação de ETC. para perto dos lábios. Assim
se constituiria o primeiro estágio de evolução do
nome da categoria cor, em LSB.
No segundo estágio, levanta-se a hipótese de
que as duas CMs (VERMELHO e ETC.) sofreram
amálgama. Justifica essa hipótese, a proximidade
de articulação dos dois itens e relativa semelhança
no parâmetro da CM de ambos.
Por hipótese, considera-se, ainda, que o ponto
de articulação da denominação de VERMELHO
é mais distintivo, icônico e dêitico que o ponto
de articulação de ETC., principalmente por que
15
Esse fenômeno de aceleração rítmica da articulação
de uma unidade lexical simples em LSB, resultante
da composição de duas outras unidades lexicais simples, permite uma análise contrastiva com fenômeno
equivalente nas línguas orais. Trata-se da alteração de
acento na composição de termos em línguas orais.
Em línguas orais, quando duas palavras se unem para
formar uma unidade lexical simples ocorre uma perda do
acento da primeira palavra constituinte da composição
e permanece o acento da sílaba tônica da segunda
palavra constituinte da unidade lexical composta. Essa
comparação foi sugerida por Heloisa Salles (UnB), em
comentários após participar da exposição na qual a
pesquisadora apresentou construção semelhante durante
divulgação de pesquisa na V Semana de Estudos Lexicais, na Universidade de Brasília, promovida pelo Centro
Lexterm na Universidade de Brasília.
69
remete à apontação de uma parte do corpo (os
lábios) que metonimicamente faz referência à cor
vermelha. Os repertórios acessíveis para pesquisa
não contêm dados suficientes para confirmar ou
refutar a seqüência reconstruída, mas é possível
entender que, num continuum desse estágio, ou,
até, no lugar dele, a CM teria passado por uma
“distensão” direta, um relaxamento na articulação
da CM, que resultaria numa das variantes em uso
atualmente (Figura 10 do anexo). Nessa hipótese,
a marca de CM, de dêitica, caso de VERMELHO,
teria flutuado entre a fixação da CM de ETC. como
ocorre na articulação do nome COR na Figura 9
(anexo), e a CM teria se distendido para toda a
mão aberta, como na Figura 10 (anexo). Também
esse exemplo parece comungar da “teoria do menor
esforço16 mencionada anteriormente (cf. FROMKLIN
& RODMAN, 1993:53).
Sinais próximos do rosto não são articulados,
normalmente, com as duas mãos. Essa formação
seria redundante e desnecessária. Esta propriedade
lingüística teria provocado o terceiro estágio de
evolução da denominação da categoria COR, em
LSB: o indivíduo lexical passaria a ser articulado
com uma única mão. Nesse estágio evolutivo da
denominação, a articulação do item resultaria em
uma CM da mão toda aberta, com movimentos
alternados dos dedos e com Movimento externo
da mão que inicialmente se posiciona diante dos
lábios e se movimenta na direção ao espaço neutro.
Então, duas das variantes tomadas como designação
para o nome da categoria cor, em LSB, são fruto da
apócope da nomeação prototípica de categorias em
LSB, constituída inicialmente, pelo vocábulo especificador da categoria “VERMELHO” acompanhado
do indicador de categoria “ETC.”
Essa análise aponta fortes indícios de que outros
nomes de categorias, totalmente opacos no presente,
podem ter percorrido caminho semelhante para a
denominação do nível superordenado da categoria
que nomeia.
Também foi possível encontrar denominações
construídas a partir da analogia a objetos do mundo
extralingüístico, cuja cor é inerente ou prototípica.
16
Um exemplo claro desse fenômeno de “relaxamento”
da CM está presente na constituição do item lexical
“e-mail” (Figura 26 do anexo), em LSB, que teve como
início, as CM
(equivalente à letra “E” do alfabeto
dactilológico) e
(equivalente à letra “M” do alfabeto dactilológico), rapidamente, modificadas para
(equivalente à letra “C” do alfabeto dactilológico) e
, respectivamente.
70
Esse grupo está designado por “nomes de cores
construídos por associação semântica”. Essa
associação semântica ocorre, especialmente, a partir de processos metonímicos, entendidos, grosso
modo, como nomes que apresentam parte pelo todo,
metafóricos, entendidos, grosso modo, como um
nome por outro e dêiticos, relacionados diretamente
com a apontação do referente no mundo real. As
hipóteses apresentadas nas análises desse grupo de
indivíduos lexicais, na maioria das vezes, não podem
ser confirmadas, pois, na sua maioria, partem das
intuições da pesquisadora em sua posição êmica.
Por isso, são suscetíveis de equívocos, embora se
constituam de hipóteses válidas. Ao observar os
indivíduos lexicais analisados nesse grupo tem-se que:
(a) Branco I (Figura 5 do anexo) parece estar diretamente relacionado à cor ‘branca’ do leite. A
representação da cor branca em Capovilla (2001)
é com uma mão mais inclinada e de leite (Figura
32 do anexo) é com a mão na posição mais
vertical. Em algumas variações diatópicas da LSB,
a articulação do leite se dá com as duas mãos,
motivada pela ordenha da fêmea bovina; Branco
II (Figura 6 do anexo) parece ser um nome
representado deiticamente pela passagem da mão
sobre a pele ‘pálida’ de pessoas definidas como
brancas. Também pode ser identificado como
extensão do nome branco registrado por Oates
(1983). Há ainda, em variação diatópica, outra
denominação para branco, a qual poder-se-ia
considerar BRANCO III. Trata-se da variedade de
branco na LSB falada pela comunidade surda de
Porto Alegre que denomina branco por meio da
apontação de um dos dentes da boca, associação
motivada, também, pela cor de um referente no
mundo real, como é o caso da associação feita
a leite para branco I.
(b) Marrom II (Figura 13 do anexo) está praticamente
representado pelo mesmo sinal de chocolate. Esse
fenômeno será chamado de metáfora antes de
se pensar em chamá-lo de sinônimo, homônimo
ou polissemia.
(c) Negro (Figura 14 do anexo) é um nome articulado
próximo à fronte e em movimento circular como
a representar uma mecha de cabelo anelado,
característica física comum a pessoas de pele
negra.
(d) Rosa (Figura 16 do anexo) é um nome articulado
na maçã do rosto. Possivelmente, esse nome está
associado à cor rosada das bochechas. Poderia
estar associado, também, ao nome rouge (hoje
conhecido por blush), que, na verdade, é palavra
oriunda do francês e designa a cor vermelha, e
não, a cor rosa. Entretanto, há muitas pessoas
que apresentam a bochecha rosada quando riem.
Portanto, pode ser considerado, até, um nome
dêitico17, apesar de sua forma não ser determinada pela apontação física, mas pela forma
da bochecha. Outra análise possível é pensá-lo
como CL que se lexicalizou como nome da cor.
(e) Vermelho (Figuras 22 e 24 do anexo) estabelece uma relação dêitica com a cor dos lábios
ou, propriamente, com a cor prototípica do
batom, a cor vermelha. A cor vermelha que
em língua portuguesa está associada ao “vermículo” (pequeno verme de coloração vermelha),
encarnado oriundo de carne, em LSB, é mais
visual e baseia-se em protótipo mais comum, a
cor dos lábios.
(f) Amarelo (Figura 1 do anexo), grosso modo, é
possível uma associação com a cor do brilho
de fios de cabelos louros. Entretanto, para se
considerar essa hipótese, seria necessário imaginar
que, no grupo onde foi criada essa associação,
houvesse, ao menos, uma pessoa, com os cabelos
louros. É comum em todos esses exemplos, o
fato de o significado metonímico passar, metaforicamente, a designar o nome da cor.
(g) A cor ‘dourada’ (Figura 30 do anexo) remete à
cor do “ouro” (Figura 33 do anexo) pelo toque
no dente ou no espaço neutro bem próximo a
ele, cuja motivação encontra-se nas obturações
dentárias, muito comuns, no passado, com
modelagem em ouro.
(h) Oates (1983) não registra a cor laranja e Capovilla
(2001) só a registra como fruta. Entretanto, nas
diversas regiões do Brasil, é comum identificar a
denominação da cor laranja pela fruta. A deno17
Os dêiticos têm a função de apontar para o contexto
situacional. A significação referencial dos dêiticos só
pode ser definida em função da situação, do contexto,
do receptor de um ato de fala. Em línguas de sinais, os
dêiticos referem-se a apontações literais dos referentes
para denominá-los ou para referir-se a eles. Há dêiticos
(a) pessoais que apontam as pessoas do discurso eu,
tu, nós, vós; (b) espaciais que referem-se a determinantes e pronomes demonstrativos, advérbios, grupos
adverbiais – aqui, cá, meu, vosso, este; (c) temporais
que se referem a advérbios de tempo, desinências
verbais temporais como amanhã, falarei; (d) sociais que
assinalam a relação hierárquica entre os participantes e
os papéis por eles assumidos, como senhora, pintora,
secretária. (In: http://portuguesessencialparaconcursos.
blogspot.com/2007/11/diticos.html).
minação representante da associação semântica
integral realiza-se com a cor Laranja (Figura 31
do anexo) que, assim como na língua portuguesa,
é homônima da fruta. O nome da cor parte da
associação semântica que se faz da cor do fruto
‘laranja’ passar à cor ‘laranja’.
Foram identificados, por fim, “nomes de cores
sem resgate da motivação”: verde II (Figura 19 do
anexo). O processo de construção não é facilmente
resgatado. Sabe-se que não foi oriundo de iniciais
emprestadas de palavra da língua oral, mas os nomes
são opacos o suficiente para dificultarem um resgate
da motivação para sua denominação.
Por fim, foi identificada composição na análise do
processo de construção dos nomes VERDE CLARO
(Figura 20 do anexo) e VERDE ESCURO (Figura 21
do anexo). O nome VERDE, ao especificar a variação
de tom, passa a pertencer ao nível subordinado
da categoria. Essa variação de tons da mesma cor
foi construída a partir do composto NOME-DACOR + CLAR@ para designar a cor verde claro,
e NOME-DA-COR + ESCUR@ para designar a cor
verde escuro. Apesar de o corpus não contemplar
outras tonalidades de cor, esse paradigma parece ser
aplicável pelos falantes de LSB para a denominação
das demais tonalidades de todas as cores, variáveis
entre tom claro e tom escuro.
Conclusão
Esse estudo, ainda que incipiente, uma vez que
identifica processos de construção de denominações
de cores em LSB, afirma a existência da COMPETÊNCIA CONSTRUCIONAL18 dos falantes de LSB,
especialmente, daqueles que a têm como primeira
língua, diante das opções que fazem ao denominar
referentes. Além de resgatar alguns dos processos de
construção dos nomes de cores em LSB, os dados
analisados mostram que os falantes preenchem,
com propriedade e organização, lacunas lexicais na
língua, sempre que necessário, a partir de processos
de construção semelhantes aos processos utilizados
pelos falantes de Línguas Orais.
Emergiu da aplicação do Modelo Silex às cores da
LSB, também, a importância e validade da análise
diacrônica para a identificação dos processos de
construção de palavras. Mais ainda, esse estudo
18
Competência Construcional é expressão cunhada por
Corbin e faz paralelo com outras expressões como as
cunhadas por Chomsky – competência lingüística – e
Hymes (1980) –competência comunicativa –.
71
mostra que o Modelo Silex sustenta uma análise
morfológica que não se prende somente a questões
estruturais, mas, também, a questões funcionais
da língua. Isso se comprova, por exemplo, nas
análises que demonstram o significado como aspecto
mais relevante na construção lexical em detrimento
da forma que tem o constituinte. A mão que, em
princípio, era tida como passiva, passa a um constituinte determinante de novas formas diante do
significado que incorpora e da função que assume
na construção de novas unidades lexicais de mesmo
campo semântico.
A análise contida nessa pesquisa não permitiu
identificar processos de conversão, nem alguns processos deformacionais como: aférese, reduplicação
de apócopes e/ou aféreses, nem sufixação avaliativa
no domínio da designação de cores em LSB. O fato
de tais processos não terem sido identificados não
significa que não existam e não sejam produtivos
em LSB. Esse fato talvez ocorra por que a análise
desenvolvida se restringiu a um corpus restrito.
Além disso, os dados foram extraídos de repertórios
lexicográficos de propósito menos funcional, cuja
natureza tenta eliminar os usos para ser mais geral
e alguns dos tipos de construção, especialmente, os
que envolvem contextos mais familiares precisam
ser extraídos de contextos reais, mais coloquiais e
mais familiares.
Por fim, o modelo que postula os três “significados” intervenientes na construção do “significado”
de uma palavra: o significado conferido pela regra
de formação de palavras, o significado herdado da
base; o significado específico do operador morfológico
(cf. CORREIA, 1999) parece ter sido confirmado
pelas análises apresentadas.
Bibliografia
1.
Capovilla, F.C. & Raphael, W.D. Dicionário Enciclopédico Ilustrado Trilíngüe
da Língua de Sinais Brasileira – LIBRAS, v. 1 e 2 São Paulo: EDUSP, 2001.
2.
Chomsky, N. Linguagem e Mente. Brasília: UnB, 1998.
3.
Corbin, D. Entre les mots possibles et les mots existants: les unités lexicales
à faible probabilité d’actualisation. In: Si lexicales. nº 1, Publication de
L’U.R.A. 382 du C.N.R.S. (SILEX) – Université de Lille III, 1997a.
4.
______. Le lexique construit. Méthodologie d’analyse. Paris: Armand Colin,
1997b.
5.
Correia, M. A denominação das qualidades: contributos para a compreensão da estrutura do léxico português. Dissertação de Doutoramento
em Lingüística Portuguesa. Lisboa, 1999.
6.
Faria-do-Nascimento, S.P. Representações Lexicais da Língua de Sinais
Brasileira. Uma Proposta Lexicográfica. Brasília: UnB / Instituto de Letras,
Departamento de Lingüística, Português e Línguas Clássicas – LIP, 2009.
7.
Ferreira, A.B.H. Novo Dicionário Aurélio da Língua Portuguesa. Dicionário
eletrônico. Versão 5.0.40. (3ª. ed., 1ª. impressão da Editora Positivo, revista
e atualizada do Aurélio Século XXI), 2004.
72
8.
Fromklin, V.; Rodman, R. Introdução à Linguagem. Tradução de Isabel
Casanova. Coimbra: Almedina, 1993.
9.
Gama, F.J. da. Iconographia dos signaes dos surdos-mudos. Rio de
Janeiro: Typographia Universal de E.; H. Laemmert, 1875.
10. Hymes, D. On Comunicative Competence. In: Pugh et al. (org). Language
and Language use, 89-104. London: The Open University Press, 1980.
11. Koch, I.V. O texto e a construção dos sentidos. Coleção: Caminhos da
Lingüística. São Paulo: Contexto, 2000.
12. Levinson, S.C. Pragmatics. London: Cambridge University press, 1983.
13. Marcuschi, L.A. Gêneros textuais: definição e funcionalidade. In: Dionísio,
A.P., Machado, A.R. e Bezerra, M.A. (orgs.). Gêneros textuais e ensino.
Rio de Janeiro: Lucerna, 2002; 20-35.
14. Mey, J.L. Pragmatics: an introduction, 2ª ed., USA: Blackwell Publishers
Inc, 1993/2001.
15. Oates, E. Linguagem das Mãos. Aparecida, São Paulo: Santuário, 1983;
1992; 325:165-167.
16. Salles, H.; Faulstich, E.; Carvalho, O.; Ramos, A.A. Ensino de língua portuguesa para surdos: caminhos para a prática pedagógica. v. 1 e
2. Secretaria de Educação Especial. – Brasília: MEC/SEESP, 2003.
Anexo – Ilustrações dos nomes de cores em LSB
AMARELO (yellow)
Figura 119
AZUL (blue)
Figura 220
BEGE I (beige I)
Figura 321
19
20
21
Figura extraída de Capovilla et al. (2001:185).
73
BEGE II (beige II)
Figura 422
BRANCO I (white I)
Figura 523
BRANCO II (white II)
Figura 624
CINZA (gray)
Figura 725
22
23
24
25
Figura
Figura
Figura
Figura
extraída
extraída
extraída
extraída
de
de
de
de
Capovilla
Capovilla
Capovilla
Capovilla
et
et
et
et
al.
al.
al.
al.
(2001:284).
(2001:314).
(2001:314).
(2001:412).
74
COR I (color I)
Figura 826
COR II (color II)
Figura 927
CORES (colors)
Figura 1028
LILÁS (lilac)
Figura 1129
26
27
28
29
Figura
Figura
Figura
Figura
extraída
extraída
extraída
extraída
de
de
de
de
Capovilla
Capovilla
Capovilla
Capovilla
et
et
et
et
al.
al.
al.
al.
(2001:466).
(2001:467).
(2001:468).
(2001:817).
75
MARROM I (brown I)
Figura 1230
MARROM II (brown II)
Figura 1331
NEGRO (black) – para pessoas
Figura 1432
PRETO (black) – para objetos
Figura 1533
30
31
32
33
Figura
Figura
Figura
Figura
extraída
extraída
extraída
extraída
de
de
de
de
Capovilla
Capovilla
Capovilla
Capovilla
et
et
et
et
al.
al.
al.
al.
(2001:873).
(2001:873).
(2001:945).
(2001:1080).
76
ROSA (pink)
Figura 1634
ROXO (purple)
Figura 1735
VERDE I (green I)
Figura 18
VERDE II (green II)
Figura 19
34
35
77
VERDE CLARO (light Green)
Figura 20
VERDE ESCURO (dark Green)
�
Figura 21
VERMELHO (red)
Figura 22
VIOLETA (Violet)
Figura 23
78
VERMELHO (red) e CORES (colors)
Figura 24
BRANCO (white), CINZENTO (gray), PRETO (black), ROXO (purple), VERDE (green) e MARROM
(brown)
Figura 25
79
E-MAIL
Figura 26
COISAS / ETC. (things; etc.)
Figura 27
VÁRIOS (several) / ETC.
Figura 28
CASTANHO (chestnut)
�
Figura 29
80
DOURADO (golden)
Figura 30
LARANJA (orange)
�
Figura 31
LEITE (milk)
Figura 32
OURO (gold)
Figura 33
81
CHOCOLATE (chocolate)
�
Figura 34
PÚRPURA/PURPÚREO (purple, Violet)
Figura 35
Adding pieces to the Portuguese Sign Language
lexicon puzzle: three pilot studies
Juntando mais peças ao puzzle do léxico da LPG:
três estudos pilotos
Ana Mineiro1*, Joana Pereira2**, Liliana Duarte3***and Isabel Morais4****
1
Instituto de Ciências da Saúde da Universidade Católica Portuguesa (ICS) – Grupo de Neurociências Cognitivas (GNC) – Fundação para a Ciência e a
Tecnologia (FCT)
2
Instituto de Ciências da Saúde da Universidade Católica Portuguesa (ICS) – Grupo de Neurociências Cognitivas (GNC)
3
Instituto de Ciências da Saúde da Universidade Católica Portuguesa (ICS) – Grupo de Neurociências Cognitivas (GNC) – Fundação para a Ciência e a
Tecnologia (FCT)
4
Instituto de Ciências da Saúde da Universidade Católica Portuguesa (ICS) – Faculdade de Letras das Universidade de Lisboa (FLUL)
Resumo
Abstract
O principal objectivo deste artigo foi apresentar as várias
peças que compõem o puzzle do léxico da Língua Gestual
Portuguesa (LGP) – a polissemia, a especialização linguística das
formas e a história dos gestos através do tempo – demonstrando
como essas peças se emolduram no mecanismo linguístico
da criatividade humana e da evolução dinâmica das línguas,
enquanto factores privilegiados de expressão cultural.
Palavras Chave: Língua Gestual Portuguesa (LGP); Léxico;
Polissemia; Terminologia; Variação Diacrónica 
A language’s lexicon is like a galaxy,
it lives in permanent expansion
for it incorporates the social and personal experiences
of the community that speaks it 1.
(Nelly Carvalho: 1989)
1. Introductory Notes
Words are the building blocks of human language
and culture. The ability to create and use words
distinguishes homo sapiens from all other animals.
In fact, nothing else is so unique and profoundly
human as the creation and usage of language.
If the ability for language is inscribed in the
modern Man’s evolution, the truth is that there are
two modalities which reveal and express both our
* [email protected]
*** [email protected]
**** [email protected])
1
Translated by the authors.
The main goal of this article is to present several pieces
of the “puzzle” concerning Portuguese Sign Language (LGP)
lexicon – polissemy, linguistic specialization of forms and the
sign’s history throughout time –, thus demonstrating how these
pieces fit into the linguistic mechanism of human creativity and
the dynamic evolution of languages, two privileged factors in
cultural expression.
Key-Words: Portuguese Sign Language (LGP); Lexicon;
Polissemy; Terminology; Diachronic Variation 
language creation ability and language usage: the
oral modality and the visual-gestural modality.
Both modalities in human languages are held
by a complex system ruled by lexicon, phonology,
morphology, syntax, semantics and pragmatics. Such
basis is present in every language in the world and
differs from one language to the other.
As far as this article goes, we will present some
of the “pieces” that compose the “lexicon puzzle”
in Portuguese Sign Language, brought about by the
most recent research work conducted in the area
of LGP. We will start by defining lexicon and the
lexicon dimensions we will be focusing on.
The word Lexicon has its origins in the Greek
term «lexicon» and, latu sensu, is a synonym of
vocabulary2. The lexicon of a language encloses its
complete inventory of words or signs as well as the
virtual possibility for creating new items (whether
signs or words). A language’s lexicon codifies the
2
In this context we will operatively perceive lexicon and
vocabulary as synonyms.
84
knowledge shared by the members of the community
that uses it, and it integrates and absorbs the new
personal and social experiences of such a community
of signers or speakers.
A given language, through the vocabulary that
connects it to the world, reflects society’s valid
culture. In the history of languages, the old forms
perish as new ones arise. Besides, the way in which
form and content relate to one another is constantly
changing. In the basis of these modifications lies an
endogenous trait of the linguistic system: creativity.
This issue is crucial because languages, the same
as signers and speakers, born and evolve, so do their
signs and words. Therefore, in their “useful life”,
lexical units that used to have a certain meaning
can come to acquire another or various others due
to the effect of polissemy, specialization of meaning
(terminology), or as a result of time and usage,
effects of the history of the language.
In this work we will account for the lexicon’s
dynamic and creative function in languages, which
renovates, increase and extinguishes words and
signs. We will do so by summing up three studies
which present these dimensions of language lexicon
concerning LGP. We will emphasize that these are
the first three research works on this domain, and
are therefore must be considered as exploratory
studies. LGP is a language without a written form and
its lexicon hasn’t yet been systematically explored,
gathered up and organized, so to provide researchers
with the qualitative and quantitative data required
for a thorough analysis of the language’s lexical
aspects. Hence, we can only raise hypotheses for
terminological, polissemic, and diachronically evolutive signs.
2. Polissemy
The concept of polissemy, which is “the association
of two or more related senses with a single linguistic
form” (Taylor 99), is a common phenomenon in
natural languages and for that it has been attended
to by researchers from diverse linguistic currents,
different branches of Linguistics. Nevertheless, not
every linguistic current has paid the same amount
of attention to polissemy. After an initial period,
centered in diachronic study, Bréal (1887) was the
first to provide a synchronic vision of polissemy
and to characterize it as a systemic phenomenon
connected to semantic change and language evolution. However, polissemy has played a rather
secondary role in linguistic studies carried out either
by structuralism or generativism.
According to structuralists, meaning analysis was
based on decomposition into semantic traits. This
process allowed them to identify each signifiersignified pair, describing it and relating it to the
contiguous meanings (through categorizing in Necessary and Sufficient Conditions, normally represented
in the form of matrixes of semantic traits). This
methodological perspective with a compositional
basis was centered on the equivalence between a
phonological form and one single meaning. It minimized perception of a phenomenon which, due to
its own nature, demanded a more global, less formal
and less compartmented approach. In a way, this
confusion between polissemy and homonymy was
recovered by generativism. Some authors referred
to this dominant paradigm during the most part of
the 20th century as the “single meaning approach”
(Cuyckens e Zawada, 2001). Hence, generativist
linguists, much more interested in the concept of
“competence” than “performance”, undertook a
semantic analysis where a lexical unit’s different
meanings were inserted in a wider, more global
meaning, described in the language’s “system”. For
this reason, describing real meanings, resulting from
usage, did not interest them.
Basically, such a lack of interest is due to both linguistic currents considering polissemy as a marginal
phenomenon, one that is never regular and systemic.
Both considered that, in the relation between form
and meaning, the predominant lexical norm was the
combination of monossemy (the “single meaning
approach”) and homonymy, that is, the formal coincidence in two lexical units which share a common
denomination but have separate meanings.
It is only when cognitive linguistics arises, throughout the 80’s and the 90’s, that polissemy starts to
play a central role in describing lexical meaning (cf.
Lakoff e Johnson, 1977; Langacker, 1991; Fauconnier,
1994; Taylor, 1995; Ungerer e Schmid, 1996, and
many others). All of these approaches shared one
same principle: that lexical units, as well as word
classes and grammatical constructions, are conceptual
categories which should be studied as a reflex of
general cognitive principles, seen as more than merely
formal linguistic phenomena. Cognitive linguistics, by
including work and methods from other academic
disciplines (philosophy of language, experimental
psychology), was better equipped to describe polissemy as a regular phenomenon of language.
Adding pieces to the Portuguese Sign Language lexicon puzzle: three pilot studies
To sum up, cognitive linguistics describes lexical
units as categories of interconnected meanings around
a prototype (Rosch, 1973), by means of semantic
associations or “family resemblances” (words well
put by Wittgenstein). Thus, the meaning of a lexical
unit would no longer have a unitary, monossemic
value, in a given profound structure, and it would
become a group of interconnected meanings by
means of cognitive processes such as metaphorization, metonymy, specialization or generalization.
Up until then, in the dominant paradigm, a polissemic word’s meanings were described as derivations of
a main meaning (usually etymologically motivated),
whereas in the constructivist paradigm, one meaning
(or several of the meanings) of a word can be more
relevant (“salient”) than the rest. Here, however,
different meanings do not derive from each other;
instead they connect to one another through the
processes we already mentioned.
In this pilot study, carried out by Mineiro et al.
(2008)3, we account for interrelations between the
different meanings of a number of LGP signs, aiming
at explaining family resemblances amidst them. We
will show that, apart from cases where the mechanism involved is metonymy stricto sensu, we can
also find others that are as yet hard to include in any
of the semantic mechanisms proposed in literature,
namely, metaphor, metonymy, specialization and
generalization.
2.1. Corpus used
The Corpus-LGP contains one hundred signs.
They were collected bearing in mind the signs
registered in Gestuário4 and in didactic5 material
used in Deaf Education, because these are the most
3
4
5
The complete original study can be found in Mineiro, A,
Duarte, L.P. Carvalho, P.V. Tebé, C. & Correia; M. “Aspectos
da Polissemia nominal em Língua Gestual Portuguesa” In:
Polissema, Vol 8, Porto, pp.37-56 , 2008.
Gestuário is a compilation of basic LGP signs, using written
contents and images, and it is similar to any oral language
dictionary. Gestuário was coordinated by António Vieira
Ferreira and Adalberto Fernandes and published by the
Secretariado Nacional para a Reabilitação e Integração das
Pessoas (National Secretariat for Rehabilitation and the Integration of People with Disabilities), in Lisbon.
The didactic materials’ author is Paulo Vaz de Carvalho, and
they were created based on the signs that are intuitively used
the most in daily communication, as well as those present in
Faria, I. H., Ferreira, J. A., Barreto, J., Martins, M., Neves, N.,
Santos, R., Vilela, S. (2002b). +LGP – Materiais de Apoio ao
Ensino da Língua Gestual Portuguesa: O Mundo. Laboratório
de Psicolinguística, FLUL. Publicação em CD-Rom, versão 1.0.
85
used sources used by LGP signers. We chose signs
that are commonly used in daily communication,
separated into the following themes: animals, fruit,
the four seasons, transports, countries and cities.
The Corpus-LGP was put together in five different
phases. In the first phase, one hundred signs were
selected from Gestuário and split into the themes
mentioned above. In the second phase, using a
written list, seven profoundly deaf subjects (with
early LGP acquisition and literate in Portuguese) were
requested to produce the signs, and such production
was captured on video. In the third phase, two
hearing fluent LGP signers and two deaf signers
selected signs that were potentially polissemic into
a sub-corpus (cf. Table 1), based on their linguistic
competence. In the fourth phase, using digitalized
images representing the elements in the sub corpus,
each informant was asked to comment freely on the
images, so that contextualized production of the
signs could be recorded and not just the isolated
units. After the several signed productions had been
analyzed, a week later, the same seven informants
were asked to repeat the fourth phase of the process,
so to try and quantify variation in the potentially
polissemic denominations.
2.2. Methods
After building the Corpus-LGP, we verified the
polissemy possibility in every signed item – using
the linguistic knowledge of two LGP fluent signers
(hearing) and two LGP signers (deaf) as reference.
Having reached conclusions on the items which
were potentially polissemic, we asked the informants
to sign the polissemic units in acceptations differing
from the initial corpus unit. We verified that certain
signs were indeed presented identically for the
several acceptations, while others were presented
with syntactic and morphophonologic variations.
In addition to such proof of variation, the truth is
certain signs sometimes appeared with variations and
other times were presented identically to the initial
form, or the “motivating” form. In order to describe
the variation observed and quantify occurrences
in the two competing forms, it was necessary for
the informants to again undergo a filmed signing
process of the polissemic units in a natural context.
This study was based on a descriptive methodology and on data observation, an approach that
intends to be data-driven and bottom-up, that is,
steered by the data and constructed in terms of
classifying it through corpus results observation.
86
2.3. Results
Bearing into mind what is generally known on
languages’ polissemy phenomenon, we verified that
LGP shows signed polissemy processes, which we
will describe in the following paragraphs.
2.3.1.Polissemy by metonymy
Traditionally, metonymy and metaphor are two
similar processes6, to the extent that they both represent systematic conceptual mapping of a sourcedomain and a target-domain. The distinction between
these two similar processes lies in the fact that
metaphor establishes similarity connections, whereas
metonymy is built on connections of contiguity. Both
processes make a decisive contribution in creating,
through extension, language polissemy.
In the case of the corpus collected, we found
metonymic polissemic signs. In 1995, Correia had
already called our attention to cases of metonymy
in LGP signs that are part of the Gestuário. From
what we know so far, metonymy is probably the
most productive process in generating polissemy in
LGP – and even in other sign languages.
In the following pairs: CAFÉ7 (drink – coffee) e
CAFÉ (place where you can have coffee – a cafe),
CEREJA (cherry) and FUNDÃO (name of a Portuguese
village); BACALHAU (codfish) and SEXTA-FEIRA
(Friday – the day when children ate codfish at the
Deaf School); and CAVALO (horse) and CARCAVELOS8 (name of a village – place where there was
a farm with many horses, at the time the sign was
crafted) we found the same sign being produced
6
7
8
In the context of the schematic network model for categorization, made popular by Langacker (1987/1991), both
metaphor in its similarity, and metonymy in its contiguity,
are presented as extension connections, opposite to schematization connections (generalization) and specification (Silva,
120 and following).
Signs are represented in capital letters because they are
glosses of LGP into written Portuguese. Despite the article’s
translation into English, these glosses are kept in Portuguese
throughout this article because it favors comprehension of
some of the linguistic aspects described.
In the first published version of this study (Mineiro et al. 2008)
the pair CAVALO and CARCAVELOS was not considered as
metonymic. It was interpreted as caused by reading deviation,
taking into account the global reading level of Deaf people.
On the occasion of the presentation of this pilot study at an
International Conference, in Brazil, May 2009, Amílcar Morais
defended that this sign should be classified as metonymic
due to its history and its crafting. Hence, we accepted this
re-reading of the polissemic process, integrating it in this
article.
to denominate both referents in each pair. In the
case of BACALHAU the sign is produced, in some
of the occurrences, using the non-dominant hand.
This happens when the sign is produced out of
context, whereas when it is produced in context, it is
identical to SEXTA-FEIRA. In the case of CAFÉ, we do
not know if its polissemy is based on a metonymic
process or in the linguistic contact there is between
the written Portuguese language and Portuguese Sign
Language. This is a possible explanation.
Other signs were found that tend to assume
metonymic polissemy. PÁSCOA (Easter) and AMÊNDOA (almond) only differed in one of the phonologic parameters: non manuals. The pair PEIXE
(fish) and TERÇA-FEIRA (Tuesday – the day when
the main course at the Deaf School was fish) also
exhibited a close proximity, thus it can come to
constitute a polissemic unit. The only difference
found between the two (PEIXE e TERÇA-FEIRA) was
reduplication9 in TERÇA-FEIRA. The trio UVA(grape)SETEMBRO(September)-PALMELA(Portuguese village)
has also shown a tendency for metonymy motivated
polissemy, and the variations in these three signs
are situated in the syntactic plan (proximal and
medial distance).
2.3.2. Polissemy by stereotype effect
LGP has several resources for the formation of
common and concrete names.
For instance, the attribution of sign names is done,
within the Deaf community, through an internal and
democratic negotiation process based on several
types of systems, as we can read in Carvalho’s
description (2006). One of the systems for attributing
sign names is the “salience effect”, meaning that
it is the selection of a physical evident trait (e.g.
big nose, small eyes, etc.) or a psychological trait
(expressivity, shyness, etc.) that leads to the sign
name’s attribution.
Another process, which is similar to this one,
refers to what happens with concrete names of
countries and cities. The name is created from one
icon that is consensually considered as representative
(stereotype) of a given location (country or city). In
the corpus collected in this study, we found several
signs for countries and cities built polissemically,
through stereotype effect. We will use this notion, in
the way Kleiber (1990) conceived it, distinguishing
9
Here we use the term reduplication, meaning the process
through which the repetition of a whole sign or the repetition
of a part of a sign occurs.
it from the notion of prototype (better specimen of
a conceptual or linguistic category). Thus, countries
and cities are named based on a sign that already
exists and which represents a stereotyped, typical
form of that same location, as we can see in the
following examples.
BRASIL (Brazil) is named after TELENOVELA (Soap
Opera) (with a variation: reduplication in BRASIL).
Both referents are named using the same sign,
only differing in the sign’s repetition, in the case
of BRASIL, a unit constructed from TELENOVELA.
The same happens with ARGENTINA and BOI (Ox),
where the reduplication happens in ARGENTINA.
Totally identical signs are IRLANDA (Ireland) and
HARPA (harp), ESCÓCIA (Scotland) and GAITA-DEFOLES (bagpipes), GUIMARÃES (a Portuguese city)
and CASTELO (castle).
With variation in one phonological parameter
(facial expression), we find the pair TERRAMOTO
(earthquake) and ITÁLIA (Italy), where the differentiating facial expression appears in the sign ITÁLIA.
With no variation, we find the pair ITÁLIA e ALGÉS
(Portuguese village), where the exact same sign is
presented for both referents in all occurrences.
2.3.3. Polissemy by linguistic contact
In the collected corpus, we think that for signs that
are already formed and stabilized, creation of new
semantic content can also happen through contact
between LGP and written Portuguese. It is a known
fact that Deaf people have difficulties in learning
to read and in written production of Portuguese.
This is described in literature on LGP and Deaf
education (Baptista: 2008), and the common “errors”
or “deviations to the written norm” are portrayed.
For this reason, we present the hypothesis10 that
reading influences the formation of the derived
acceptation in the pre-existent sign.
The process that we present here holds two interesting variants, linguistic contact with no reading
deviation and deviant linguistic contact. The fact
that we consider the process we present here as
polissemic instead of homonymic comes from having
operatively limited the notion of polissemy and
using this concept whenever there is an intentio10
To think that the written form in Portuguese motivates the
creation of LGP signs, through polissemy, is a plausible
interpretation but yet to be proved. In order to understand
if this is a recurring process, we would have to verify this
hypothesis in a statistical study that would allow us to reach
a trustworthy conclusion.
87
nal and rational connection between the various
acceptations of a linguistic item (sign/word). In
this case, we consider that there is a relation of
linguistic contact11 between two languages in one
community – Portuguese is the “written” language
of Portuguese deaf people and it promotes, through
the reading channel, an interpretation of two items
as related to one single form.
A possible example, which illustrates the creation
of an acceptation within a pre-existent sign, due
to a reading deviation 12 in Portuguese, is the pair
BRISTOL and PISTOL(A) (gun). In this pair we see a
similarity in writing, between the groups of consonants and vowels used in BRISTOL and PISTOL, which
can lead to the creation of an acceptation based
on the pre-existent signifier-signified relationship.
There are cases, such as the pairs PERU (animal)
and PERU (country), and CAFÉ (coffee – the drink)
and CAFÉ (location where one drinks coffee), where
the similarity of the two signs in each pair seems
to be anchored in a reading process of written
Portuguese, with no deviations. As we said before,
it is impossible to perceive whether CAFÉ (the
drink) and CAFÉ (location) are polissemic forms by
metonymy within LGP, or if reading in Portuguese
has influenced naming these two referents using
the same sign.
2.3.4. Polissemy by imagetic synonymy
One of the processes for “recycling” signs for
inexistent referents was what we thought to have
found through visual image. This process seems
to be particularly interesting, for it is believed that
vision is one of the highly developed senses in
Deaf People. Consequently, processing the “image
world” will linguistically be an operative process and,
namely, an LGP process. The signs’ morphological
composition processes are most of all visually motivated, generally referential (indicating indirectly the
parts of the body or pronouns), iconic (delineated
representation of the object or using hand shape
to represent the object itself) and metaphorical and
11
12
We must enphasize that the word creation in oral languages,
through linguistic loans (linguistic contact), is also done
through the unit of origin’s “linguistic deformation” (e.g
abajur, quivi, líder, among others). On this matter, one can
consult works on European Portuguese, such as the ones by
Rebello de Andrade and Lavouras Lopes (2003).
We think that this process is common to other sign languages,
namely Brazilian Sign Language (LIBRAS) and British Sign
Language (BSL). This hypothesis can be further looked into
in future studies.
88
metonymical (cf. Hub-Faria et al. 2001: 87-98). It
seems to us that the “imagetic synonymy” process
proposed here fits in LGP’s tendency to create signs
that are visually motivated.
In this case, we found pairs concerning brand
names where their symbol is a sign that already
exists, such as in ELEFANTE (elephant) and JUMBO13
(supermarket brand); ESTRELA (star) and AMADORA
(location in Lisbon) (due to the influence of there
being a Football Club called “Estrela da Amadora”).
Imagetic synonymy (the representative symbol and
its referent), in these cases, leads to the attribution
of a sign that is identical to its source, resulting in
a similar form with several meanings (in the cases
above, ELEFANTE equals JUMBO, and ESTRELA
equals AMADORA).
2.4. Description of potentially polissemic sign
variation
Within natural languages, variation is systemic.
LGP’s youth can be a promoter of such an internal
facet in languages. We can find this in literature,
described for example by Henriques (2006), concerning nominal variation in the “história da rã”
(the frog story).
An interesting issue was raised during data collection; we acknowledged that in such a young
language as this one, there are indeed potentially
polissemic linguistic forms. Still, in this context,
they haven’t yet reached a state where they are
completely stabilized.
In that case, if some of the collected signs are
clearly polissemic linguistic forms, that is, one same
sign for several co-relatable meanings, other signs
compete with each other to become enveloped in
this phenomenon.
We think that the “linguistic economy” factor
that Aristóteles14 spoke of, referring to the reasons
for which polissemy exists, can make a decisive
contribution for polissemic forms’ “natural selection”
do be done to the detriment of its variants.
13
14
Jumbo is a supermarket chain which symbol is an elephant.
Aristóteles finds a correct reason – linguistic economy, meaning
the recycling ability of linguistic matter in the face of new
referential stimulli (objects, concepts) – to explain polissemy,
when he states that:
Names exist in a limited number, as well as the plurality of
enunciates, whereas this are finite. It is therefore inevitable
that the same enunciate and that one same single word means
several things. (Aristóteles, Elencos Sofísticos, 165a 10-13, apud
Silva, 16 – translated by the author).
There were signs which were undoubtedly polissemic (100% of occurrences in the collected corpus),
while other signs presented some type of variation,
whether concerning syntactic parameters – the same
sign being produced in different distances in the
syntactic space (proximal, medial and distal); whether
relating to morphophonology (such as the sign’s
reduplication or a noticeable difference in facial
expression).
The truth is that signs which presented derivations
in relation to their original form (proofing to be
according to our opinion, competing variants) do
not seem to be yet stable in their differentiated
form and have shown occurrences where the two
forms are identical (cf. Table 2). Only LGP’s future
history will be able to shed some light whether on
one form prevailing over the other, or both forms
remaining in usage.
To sum up, the signs that were presented in every
occurrence with one same form were:
BRISTOL – PISTOL(A); CAVALO – CARCAVELOS;
CEREJA – FUNDÃO; CASTELO – GUIMARÃES; CAFÉ
(location) – CAFÉ (drink); ELEFANTE – JUMBO;
ESTRELA – AMADORA; HARPA – IRLANDA; ITÁLIA – ALGÉS; GAITA-DE-FOLES – ESCÓCIA; PERU
(animal) – PERU (country).
The signs that presented variation were:
BACALHAU – SEXTA-FEIRA; BOI – ARGENTINA;
ESTRELA – AMADORA; UVA – PALMELA – SETEMBRO;
TERRAMOTO – ITÁLIA; PÁSCOA – AMÊNDOA; PEIXE
– TERÇA-FEIRA; TELENOVELA – BRASIL.
2.5. Concluding Remarks
This work presented here is one first systematic
approach of this theme. We recognize this study’s
“embrionary” nature, but we think it already presents
some relevant aspects. It is a work of observation
and data description on the polissemy phenomenon,
an analysis that as far as we know has not yet been
conducted concerning LGP. It is an attempt to classify
processes that are subjacent to polissemy in LGP,
processes that seem to be a feature of this language.
We believe that the causality hypothesis we have
enunciated connecting the polissemic form to the
polissemy’s origin, particularly in the case of sign
formation through reading words in the Portuguese
language, need to be developed in future research.
3. Specialization of meaning: terminology
in LGP
A language’s lexicon progressively expands in
new forms, many of which are built from forms that
already exist. These new forms result from linguistic
creativity as well as human creativity in other fields.
Neologisms created in science, technology or art
provide us with a linguistic passage to have access to
new concepts, therefore accompanying the evolution
of societies. We will see that these new linguistic
forms are words or signs which contain specialized
meaning; they are used in certain linguistic communicative contexts (e.g. science, technology, arts)
and obey to the same linguistic creation rules than
common lexical items (although there are preferential
processes for term formation described for languages
in the oral modality).
Nowadays, there is an enourmous gap between
the interest in LGP terminology and elaboration of
terminological work. Studies such as Terminologia da
Análise Linguística da Língua Gestual (Terminology
of Linguistic Analysis of Sign Language) (Prata;1994),
Emergência de uma Terminologia Linguística em
Língua Gestual Portuguesa (the Emmergence of a
Linguistic Terminology in portuguese Sign Language
(Delgado-Martins; 1998), and digital thematic dictionaries such as A Casa (The House) (Faria et al;2001),
O Corpo (The Body) (Faria et al; 2002) and O Mundo
(The World) (Faria et al;2002) are some of the very
few works done in this field.
The lack of specialized signs is felt not only by
LGP interpreters but also by Deaf students who now
have access to higher levels of education, and thus
need specialized signs to refer to concepts that are
specific of their training area.
The goal of this pilot study undertaken by Duarte
et al. (2007)15 was to build an inventory of LGP
signs collected in a classroom, and to analyze the
formation processed used by Deaf students. Due
to the impossibility of exploring a wider corpus,
on area was chosen for scrutiny: Natural Sciences,
a subject present in the “Ensino Básico” (Basic
Education) Curricula .
15
The complete study was originally published in Duarte, L.,
Mineiro, A “Terminologia em Língua Gestual Portuguesa:
Uma necessidade para a tradução? Processos de formação de
Gestos em Ciências Naturais” In: Encontro Comemorativo dos
50 anos do Centro de Linguística da Universidade de Lisboa
(CLUL), Lisboa, 2007. http://www.clul.ul.pt/artigos.php.
89
3.1. Method
In a first phase, contact was established with every
“Unidades de Apoio Educativo a Alunos Surdos”
(Units for Educational Support of Deaf Students) in
the Portugal – more commonly known as UAEAS.
They were a total of thirty two, according to the
information provided by DGIDC – “Direcção-Geral de
Inovação e de Desenvolvimento Curricular” (General
Board for Innovation and Curricular Development).
This information was gathered at the UAEAS National
Meeting, in December 2005.
In a second phase, criteria we established for
selecting schools. These included all of the important
variables to take into account in this sort of study:
– Schools which included one UAEAS;
– With one or more LGP interpreters;
– With one or more LGP teachers;
– With classes that had the largest number possible of signing deaf students, in the subject of
Natural Sciences/ Sciences of Nature, Biology/
Geology and Geology, with an LGP interpreter
in the classroom;
The defined conditions reduced our universe of
study because not every school that could participate
in this study showed interest to do so. Therefore,
our universe was cut down to 8 students, from the
7th and 8th grades, with ages between 13 and 19
years (in Coimbra).
In a third phase, video recordings were made at
the only school that filled the requirements set by
us in the beginning. These footages carried out in
6 classes, with a duration of 45 minutes each.
In a fourth phase, researchers proceeded to sign
collection and analysis, based on the terminological
lexicon the students produced in the classroom.
Finally, on a fifth phase, the results obtained (the
terms registered in the footages) were analyzed, with
the assistance of a deaf native LGP signer.
3.2. Corpus collected
Atmosfera (atmosphere); CFC (clorofluorcarbono) (chlorophluorocarbons); Chuvas ácidas
(Acid rain); Epicentro (Epicenter); Hipocentro
(Hipocenter); Maremoto (Seaquake); Marés negras
(Black Tides); Ozono (Ozone); Poluição (Polution); Sismo (Seism); Tsunami; Raios Ultravioleta
(Ultraviolet Rays).
90
3.3. Results
The analysis of the results obtained focused on
two different parameters: a qualitative parameter and
a quantitative parameter. In both cases the analysis
was limited to a descriptive observation of the data.
In the future, we hope to widen the corpus so to
obtain results that can allow us to withdraw broader
conclusions on the resourcefulness and creativity of
the formation processes found.
Following the qualitative parameter, the obtained
signs were classified according to the type of formation they presented whereas in the quantitative
parameter, the signs were analyzed according to
descriptive statistics data.
The methodology used in our corpus qualitative
analysis was similar to the one presented in the study
of Raquel Delgado Martins (1998). We considered
that quantitative analysis can be an interesting option
if this study is broadened in the future, which is
why we decided to include it in our work.
3.3.1. Qualitative descriptive analysis
The sign formation processes produced by the
students were scarce since some LGP sign formation processes are not represented here. Formation
processes such as derivation, initialization16,
paraphrasis or external loan were not found.
On the one hand, we think this is due to the
reduced dimension of the collected corpus and, on
the other hand, to its thematic limitation because
it is confined to one specific area of knowledge.
The sign formation processes found were:
(1) Dactilology: process through which the term
is shaped by manual configurations corresponding
to the manual alphabet. This process involves some
knowledge of written Portuguese, which might not
happen in the case of deaf people with a low
educational level.
In the universe of this study, signs formed by
dactilology were:
A-T-M-O-S-F-E-R-A; C-F-C; E-P-I-C-E-N-T-R-O; H-IP-O-C-E-N-T-R-O; M-A-R-E-M-O-T-O; O-Z-O-N-O;
T-S-U-N-A-M-I; R-A-I-O-S U-L-T-R-A-V-I-O-L-E-T-A.
16
This term, already used in various studies in LIBRAS (Brasilian Sign Language), was used by us in alternative to the
syntagmatic term “initial dactilologic configuration”, proposed
by Amaral, Coutinho and Delgado Martins (1994) because
we think it to be more economical from a linguistic point
of view and conceptually more transparent.
(2) Internal Loan: process of reusing an existent
sign in the common LGP lexicon, conferring a specific
meaning of an area of specialty to it.
The sign formed through internal loan was: SEISM.
In the case of Seism, students used the sign for
Earthquake to refer to Seism.
Although there seems to be a dim line dividing
the concepts seism and earthquake, there are indeed
differences. Concerning the concept of seism, we
can say that it refers to a quake or vibration that
occurs inside the Earth and results from breaking
elastic tensions. However an earthquake refers to
land shifting. The same process occurs for the term
seaquake, which means the existence of underwater
movement.
We can therefore consider, in this case, that there
is a semantic hierarchical connection, where seism
appears as a hyperonim of earthquake and seaquake.
(3) Composition (sign + sign and sign + sign
+ sign): process where the concept is expressed
through the junction of existent signs, with no
morpheme loss.
The composed signs were:
CHUVA ÁCIDA (Acid rain); MARÉS NEGRAS (Back
Tides) and POLUIÇÃO (Pollution)
In the case of acid rain, the students used the
signs CHUVA (rain) and SUJO (dirty) aggregated
sequentially. This composition seems to derive
from a visual linguistic motivation, since proximity
between the object and its denomination is perceivable. It is also important to mention that the
concepts were transmitted by the subject’s teacher
through strategies that included explanation, usage
of pantomimic gesture, the image referring to the
object, and the presentation of negative and positive
effects of the concepts. Hence, for acid rain students
used CHUVA (rain) + SUJO (dirty) (because of the
visible degradation caused to monuments) and
for black tides BARCO (boat) + PARTIR (depart) +
DERRAMAR (spill) (because of the visual image of
the oil tankers).
For pollution, students opted by the combination
of the signs SUJO (filthy) + ESPALHAR (spread). This
association allows us to say that these two signs,
clustered together, transmit the perfect understanding
of the concept, since pollution is related to filth
and is not something that is fixed in one particular
location.
3.3.2. Quantitative descriptive analysis
So to better visualize statistical distribution of the
formation processes in the collected corpus, we
present the following graphic:
8%
25%
67%
Dactylology
Internal Loan Sign
Composition
As we can see, 67% of the signs were produced
through dactylology. We also observe that composition of signs obtained a 25% result whereas the
internal loan process attained 8% of the occurrences.
These results show a clear predominance of dactilology as a privileged sign formation process in
this area of knowledge. The reasons for this result
can be connected to the fact that, when the footage
was being captured, students were being exposed
to these concepts for the first time. So, there might
not have been “enough time” to create processes
that are endogenous to LGP and are alternatives to
dactylology. The dactilology sign formation process
is natural in a school community because it involves
strategies of identification written Portuguese. If the
target population was made of Deaf people with a
low level of education or illiterate subjects, results
could have shifted and LGP neology would manifest
itself in other formation processes.
Sign composition seems to be an alternative to the
previous process with some representativity. This can
be due to available linguistic material (existent signs)
being used to create new denominations, which
demonstrates LGP’s ability to recruit economical
morphological processes. Like any other language
whether of an oral or signed modality, LGP has a
set of rules that allows it to generate an infinite
number of enunciation and signs.
The internal loan process presents a result of 8%.
We can consider the collected corpus as an early
stage of our study, which scope should be broadened
not only within this same thematic area, but also in
other curricular areas that interest the Deaf community and every professional that works with LGP.
91
We come to the conclusion that the formation
processes used in this study are parallel to those
found in oral languages. From a linguistic point
of view, this reinforces what biolinguistic studies
have been demonstrating: sign languages are true
linguistic systems that only differ from oral languages
in the modality used for expressing and perceiving
them.
Even though dactilology was the sign formation
process that was the most used by our subjects, we
think that happened because LGP signers, in the
classroom context, hadn’t had time to completely
assimilate the concepts so that more specialized signs
would arise naturally, using formation processes
which were endogenous to LGP’s morphological
system. This result replicates the one described in
Amaral, Coutinho & Delgado Matins (1994).
We believe that if the students had been filmed
during a longer period in time, alternatives dactilological signs would have been found in the
classroom context. The dactylology process can be
compared to what happens when, in the Portuguese
language, we first receive a term belonging to a
foreign language. Such a term is not immediately
integrated. Firstly, it is used just as it arrived to us: in
the form of the language of origin. The integration
phase and assimilation occurs latter on.
4. The history of signs in the semantic
field of family
«The history of a word is the history of its culture
and structure;
Both aspects should really be described in
relation to each other,
as if they were two sides of a same coin17»
(Helmut Ludtke, 1968)
Although Portuguese Sign Language (LGP) is
more than two-hundred-years-old, the truth is it was
only formally recognized in 1997, having become
a legally acknowledge language in the Portuguese
Constitution.
Signs used in daily communication are easily
changeable, ever evolving into new forms, and this
often steals away the historical background of the
signs themselves. Processes of linguistic economy
tend sometimes to simplify the original signs, transforming the lexical cluster of sign languages.
17
Translated by the authors.
92
In this exploratory study, conducted by Pereira
et al (in press)18 we intend to retrace the historical
path and the etymology of a semantic field: family.
4.1. Methods and Materials
Data collection was obtained with the participation
of 15 Deaf signers, LGP natives which were divided
into three age groups: (i) more than 10 years of
age, (ii) more than 25 years of age, (iii) and more
than 45 years of age. Subjects were selected using
high LGP fluency as a main criteria and the research
team tried to make the sample as heterogeneous
as possible in terms of each of the participants’
language acquisition and educational background.
Our sign corpus included 32 family ties, for which
we intended to search etymology explanations. The
concepts used were: mother, father, grandmother,
grandfather, son, daughter, uncle, aunt, godfather,
godmother, godson, goddaughter, brother, sister, cousin (masculine), cousin (feminine), brother-in-law,
sister-in-law, father-in-law, mother-in-law, stepson,
stepdaughter, stepfather, stepmother, nephew, niece,
daughter-in-law, son-in-law, great-grandfather,
great-grandmother, grandson and granddaughter.
Subjects were asked to explain the origin of
each sign, using a written list of the words above.
Proposals of origin for the signs were registered
and accounted for. The most mentioned etymology proposal was registered, as well as its level of
cohesion, depending on the number of participants
who had suggested it; Weak [1 to 5]; Medium [5 to
10]; and High [10 to 15]. Within these categories we
also identified others: Weak– [1 to 2]; Weak* = 3;
Weak+ [4 to 5]; Medium– [6 to 7]; Medium* = 8;
Medium+ [9 to 10]; High– [11 to 12]; High* = 13;
High+ [14 to 15]. Where no answer was obtained
results were named Null.
4.2. Results
According to the most mentioned etymology proposal by our informants, a list of proposal for the
etymology of each sign was elaborated (see Table
3). We intend to apply such a list in further research
using a wider sample of Portuguese Deaf participants,
18
The complete study on diachrony in the semantic field of
family is being published in Pereira, J. Morais, I.; Duarte, L.
Morais, A. & Ana Mineiro “Diachronic Variation in Portuguese
Sign Language” In: Proceedings of the 1st Symposium in
Applied Sign Linguistics, University of Bristol, Centre for Deaf
Studies, Bristol: UK.
so to determine etymology more accurately in this
semantic field. However, and although our informant
group was small, some aspects were noticed and
are worthy of reflection. These issues pointed in
six directions, which can constitute future research
hypothesis, to be confirmed through upcoming
studies.
1 – Most easily identified signs
The signs which origins were most easily identified (or suggested) by the informants were PAI
(father), MÃE (mother), BISAVÔ (grandfather),
BISAVÓ (grandmother), MADRINHA (godmother),
PADRINHO (godfather), AFILHADO (godson), AFILHADA (goddaughter), MADRASTA (stepmother)
and PADRASTO (stepfather). These results might
have to do with the fact that all these signs refer
to direct family members, people that are close to
our informants. Perhaps they end up being signs
that are more used in signed conversations and,
being often part of dialogues and discussions, their
origins become more probable to be talked about
and transmitted to others.
2 – Gender similarity
From the 32 signs analyzed, made of 16 gender
pairs (e.g. father/mother, son/daughter, etc.), 12 pairs
(24 signs) matched in terms of similar justifications
given in the group. We only registered differing
etymology proposals in the pairs PAI/MÃE (father/
mother); and ENTEADO/ENTEADA (stepfather/stepmother). It is interesting to observe that gender
matches even happened in the pair PADRASTO (stepfather) (PAI-father+SEGUNDO-second)/MADRASTA
(stepmother) (MÃE-mother+SEGUNDO-second). In
this case gender is not marked adding the sign
FEMININO (feminine) in the beginning but using
the sign MÃE (mother) instead.
3 – Knowledge vs inventive ability
This study’s participants have probably made use
of their inventive abilities in order to explain the
signs’ origin. Since there is no written record of
old LGP signs, we accepted that the justifications
provided might come from knowledge transmition
from the older to the youngest generations. Nevertheless, we must also consider that the etymology
proposals collected can also have been mere explanations given in the moment of questioning, based
on the presented sign’s visual motivations and the
informant’s sensibility as an LGP signer and member
of the Deaf community.
4 – Teaching the language in study
In group (ii), an interesting connection was
observed: LGP teachers gave more sign etymology
proposals than the rest of the informants. Teaching
the language in study might be a factor for having
a deeper knowledge on the language’s terms etymology.
5 – Historical factor
People in group (iii) (over 45) were educated
under the oralist system and during the period
Salazar’s dictatorship was in power. At the time,
before the Revolution of the Carnations (25th April
1974), divorces were scarce and, therefore, signs
which express concepts related to the reality of
divorce (stepfather, stepfather, stepdaughter, stepson,
children of the stepfather or stepmother) are always
used recurring to dactylology.
This occurs in contrast with what happens in
group (i), where the youngest Deaf participants are.
In the Portuguese educational system these students
already have available Deaf models (Deaf native LGP
signers as their teachers), with whom they create
a specific dynamic where their language’s past is
absorbed and its future is shaped. In this process
of language creation and evolution, language terms
appear which are already adapted to the social
conditions of the present time, such as signs of
family ties spurring form divorces. In fact, the most
part of these youngsters’ parents are divorced.
6 – Co-relation bewteen age group and metalinguistic conscience
The group where more participants provided less
than 20 answers with no justification, that is, where
they just said they did not know the sign’s origin,
was group (ii) (4 informants), followed by group (i)
(3 informants), and at last group (iii) (1 participant).
This seems to indicate that in the sample we studied,
children and young adults reveal more knowledge
on the sign’s origins than elder participants. In a
way, this contradicts our initial expectations, since
we believed the older the participant the more
knowledge he/she would have on etymology. This
observation, which is opposite to the research team’s
expectations, can probably be justified by the fact
that nowadays LGP is studied and formally taught,
which did not happen in the old days. Thus, a
metalinguistic conscience arose from acknowledging
and teaching the language.
93
This exploratory study led us to congregate a list
of etymological proposals for signs in the semantic
category of family, to be tested against a wider
population of participants in the future. We also came
up with a few hypotheses that we intend to explore
in a near future, namely the correlation between
the age group and sign etymological conscience,
as well as the didactic interest of Deaf teachers in
etymological findings.
5. Final remarks
In this work, which includes three previously
published pilot studies, we intend to account for
the fundamentally dynamic features of LGP lexicon.
By presenting the effects of polissemy or meaning
extension, of the form’s specialization into naming
specialized concepts, and the history of common use
signs in three age groups, we provided the reader
with a few pieces of the complex puzzle which lies
underneath language lexicon, particularly that of LGP.
We therefore wish to have contributed and continue making contributions in LGP lexicon studies,
using the hypothesis raised by these preliminary
studies.
Such future studies can come to contribute to a
better knowledge on LGP and the creation of tools
(e.g. vocabularies, dictionaries, grammars) which
will be useful for LGP’s development as a teaching,
cultural and scientific language.
6. References
1.
Amaral, M. A., A. Coutinho e M. R. Delgado Martins. Para Uma Gramática
da Língua Gestual Portuguesa. Lisboa: Caminho, 1994
2.
Baptista, J. A. N. Os Surdos na Escola. A Exclusão pela Inclusão. Tese de
Doutoramento. Universidade Católica Portuguesa. Viseu, 2007.
3.
Bréal, M. “The history of Words”. The beginnings of semantics. Essays, Lectures
and Reviews. Ed. Wolf, G. London: Duckworth, 1887.
4.
Carvalho, P. V. Contribuição para o estudo da formação de atribuição dos
nomes gestuais nas comunidades de surdos em Portugal. Diss. de Mestrado.
Faculdade de Ciências Sociais e Humanas da Universidade Nova de
Lisboa. Lisboa, 2006 (inédita).
5.
Carvalho, N. Empréstimos Linguísticos, São Paulo, Ed. Ática, 1989.
6.
Correia, M. “O léxico na Economia da Língua”. Ciência da Informação 3
(1995): 299-306.
7.
Cuyckens, H. e B. Zawada, eds. Polysemy in Cognitive Linguistics. Amsterdam:
John Benjamins, 2001.
8.
Duarte, L., Mineiro, A “Terminologia em Língua Gestual Portuguesa: Uma
necessidade para a tradução? Processos de formação de Gestos em
Ciências Naturais” In: Encontro Comemorativo dos 50 anos do Centro
de Linguística da Universidade de Lisboa (CLUL), Lisboa, 2007. http://
www.clul.ul.pt/artigos.php.
94
9.
Henriques, L. Para uma descrição da variação nominal em LGP. MA Diss.
F.L.U.L. Lisboa, 2007 (n/p).
10. Hub Faria, I, L. Henriques, M. Martins e O. Monteiro. “Predicados de movimento em Língua Gestual Portuguesa”. Revista Polifonia. Ed. Colibri, nº
4 : Lisboa (2001): 87-98.
11. Fauconnier, G. Mental spaces. Cambridge: Cambridge University Press, 1994.
12. Kleiber, G. La sémantique du prototype. Catégories et sens lexical. Paris:
Presses Universitaires de France, 1990.
13. Klima, E. e U. Bellugi. The Signs of Language. Harvard, U.S.A.: Harvard
University Press, 1975.
14. Langacker, R. Foundations of Cognitive Grammar, Vol. 1, Theoretical Prerequisites. Stanford: Stanford University Press, 1987.
15. Langacker, R. Foundations of Cognitive Grammar, Vol. 2, Descriptive Application. Stanford: Stanford University Press, 1991.
16. Lakoff, G. Women, Fire and Dangerous Things: What Categories Reveal about
the Mind. Chicago: The University of Chicago Press, 1987.
17. Lakoff, G. e Mark Johnson. Metaphors we live by. Chicago: The University
of Chicago Press, 1980.
18. Lyons, J. Semantics. Cambridge: Cambridge University Press, 1977.
19. Ludtke, H. Historia del lexico románico, versão espanhola de Hernandez,
M. Madrid, Credos, 1968.
20. Mineiro, A, Duarte, L.P. Carvalho, P.V. Tebé, C. & Correia; M. “Aspectos da
Polissemia nominal em Língua Gestual Portuguesa” In: Polissema, Vol 8,
Porto, pp.37-56 , 2008.
21. Martins, Maria Raquel Delgado (1998) “Emergência de uma terminologia
linguística em Língua Gestual Portuguesa” In: Actas do XIII Encontro
Nacional da Associação Portuguesa de Linguística, Lisboa: APL / Colibri.
22. Palmer, F. Semantics. A New Outline. Cambridge: Cambridge University
Press, 1976.
23. Pereira,J. Morais, I Duarte, I Morais, A. & Ana Mineiro “Diachronic Variation
in Portuguese Sign Language” In: Proceedings of the 1st Symposium in
Applied Sign Linguistics, University of Bristol, Centre for Deaf Studies,
Bristol: UK.
24. Petitto, L. A. e P. F. Marenttete. “Babbling in the Manual Mode: evidence
for the ontogeny of language”. Science , vol 251, (1991): 1493-1496
25. Poizner, H., E. S. Klima e U. Bellugi. What the Hands Reveal about the Brain.
Cambridge, MA: MIT Press, 1987.
26. Rebello de Andrade, A. As palavras importadas no léxico da decoração.
Diss. de Mestrado. Faculdade de Letras da Universidade de Lisboa. Lisboa,
1996 (inédita).
27. Rebello de Andrade, A e A. Lavouras Lopes. “O tratamento dos estrangeirismos nas duas últimas edições da Porto Editora”. Revista de Lexicografía,
IX (2003): 7-28
28. Rosch, E. “Natural Categories”. Cognitive Psychology. Vol 4, Academic Press
Inc (1973): 328-350.
29. Silva, A. Soares da. O Mundo dos Sentidos em Português: Polissemia, Semântica e Cognição. Coimbra: Almedina, 2006.
30. Stokoe, W. Sign language structure. An Outline of the visual communication
system of the American deaf. Studies in Linguistics. Occasional Papers 8.
Buffalo, N.Y.: University of Buffalo, 1960.
31. Taylor, J. Linguistic Categorization: Prototypes in Linguistic Theory. Oxford:
Clarendon Press, 1995.
32. Ungerer, F. e H. J. Schmid. An Introduction to Cognitive Linguistics. London:
Longman, 1996.
33. Wittgenstein, L. Philosophical Investigations. Oxford: Blackwell, 1953.
Table 1 – A few polissemic or potentially polissemic signs
BACALHAU/ SEXTA-FEIRA
(codfish/Friday
CAVALO/ CARCAVELOS
(horse/Carcavelos)
BOI/ ARGENTINA
ox/Argentina
CASTELO/ GUIMARÃES
(castle/Guimarães)
ESTRELA/ AMADORA
(Star / Amadora)
PÁSCOA/ AMÊNDOA
(Easter/almond)
UVAS/ SETEMBRO/ PALMELA
(grapes/September/Palmela)
GAITA-DE-FOLES/ ESCÓCIA
(bagpipes/ Scotland)
PEIXE/ TERÇA-FEIRA
(fish/Tuesday)
BRISTOL/ PISTOLA
(Bristol/pistol)
CAFÉ (bebida) / CAFÉ (local)
(coffee-drink/cafe-location)
CEREJA/ FUNDÃO
(cherry/Fundão)
ELEFANTE/JUMBO
(elephant/Jumbo)
HARPA/IRLANDA
(harp/Ireland)
PERÚ (animal)/ PERU (pais)
(turkey-animal/Peru-country)
95
96
Table 2 – Frequency of competing occurences1
Signs
1
2
Total No. of ocorrences
Production of the same sign
ALGÉS
7
7
ITÁLIA (Italy)
8
8
TERRAMOTO (earthquake)
6
5
1
AMÊNDOA (almond)
9
6
3
PÁSCOA (Easter)
7
6
1
BACALHAU (codfish)
7
7
SEXTA-FEIRA (Friday)
7
7
BOI (ox)
9
7
ARGENTINA
8
8
BRASIL (Brazil)
9
6
TELENOVELA (soap opera)
7
7
CAFÉ (coffee-drink)
7
7
CAFÉ (cafe-location)
7
7
CAVALO (horse)
8
8
CARCAVELOS
8
8
CASTELO (castle)
7
7
GUIMARÃES
7
7
CEREJAS (cherry)
9
9
FUNDÃO
9
9
ESTRELA (star)
3
2
1
AMADORA
14
8
6
ELEFANTE (elephant)
3
3
JUMBO
7
7
GAITA-DE-FOLES (bagpipes)
5
5
ESCÓCIA (Scotland)
7
7
HARPA (harp)
4
4
IRLANDA (Ireland)
9
9
PEIXE (fish)
10
7
TERÇA-FEIRA (Tuesday)
8
8
PERÚ (turkey-animal)
10
10
PERÚ (Peru-country)
6
6
PISTOLA (pistol)
7
7
BRISTOL
7
7
UVAS (grapes)
8
8
SETEMBRO (September)
6
6
PALMELA
8
6
In this column are the signs which presented more occurrences comparing to the column “Variation”.
Signs produced with phonological trait variation which does not compromise its meaning.
Variation2
2
3
3
2
97
Table 3 – Most enunciated etymological proposals for signs in the semantic category “Family”
SIGN
COHESION
ETYMOLOCICAL PROPOSAL
PAI (father)
Medium+
Association with the sign HOMEM (man) e BIGODE (moustache), both very
similar to the sign PAI (father).
MÃE (mother)
Weak+
Related to the old of sons and daughters kissing their mother’s hands;
Protectiveness of the mothers towards their children.
Weak-
Dactilology influence: “A”, first letter in “avô” (grandfather) and initial handshape
of the sign;
Association of the idea of “grandfather” with the concept of “old”. The location
at the chin comes from the location of the sign VELHO (old).
AVÓ (grandmother)
Weak-
Dactilology influence: “A”, first letter in “avó” (grandfather) and initial handshape
of the sign;
Association of the idea of “grandmother” with the concept of “old”. The location
at the chin comes from the location of the sign VELHA (old).
BISAVÔ (greatgrandfather)
(AVÔ+SEGUNDO)
(grandfather+second)
Medium*
It is the second grandfather.
BISAVÓ (greatgrandmother)
(AVÓ+SEGUNDO)
(grandmother+second)
Medium*
It is the second grandmother.
Weak-
Association with the sign MÃE (mother) and the concept mother:
– “possession of the mother”;
– ”that is after the mother in the genealogic tree”
– “that is born from the mother”.
Weak-
Association with the sign MÃE (mother) and the concept mother:
– “possession of the mother”;
– ”that is after the mother in the genealogic tree”
– “that is born from the mother”.
Weak+
“someone that lives as an equal” to a brother (the sign for equal is the same as
for brother in LGP);
Someone who “grows up with you, and is always near you”
Someone who has the “same blood as you”.
IRMÃ (sister)
(FEMININO+IRMÃ)
(feminine+sister)
Weak+
“someone that lives as an equal” to a brother (the sign for equal is the same as
for brother in LGP);
Someone who “grows up with you, and is always near you”
Someone who has the “same blood as you”.
TIO (uncle)
Weak*
Dactilology origin: T-I-O.
TIA (aunt)
(FEMININO+TIO)
(feminine+uncle)
Weak*
Dactilology origin: T-I-A.
PRIMO (cousin-male)
Weak-
Visual triangle that is formed in the genealogic tree and illustrates the relation
between two cousins or two families.
PRIMA(cousin-female)
(FEMININO+PRIMO)
(female+cousin)
Weak-
Visual triangle that is formed in the genealogic tree and illustrates the relation
between two cousins or two families.
Weak*
Old family ties between nephews/nieces and uncles/aunts. The latter were also
godfather/godmother to the first. These people were connected not only by a
consanguinity issue but also by the ties imposed by a religious ceremony: baptism.
Hence, the location and movement of SOBRINHO (nephew) is the same as
the location and movement in PADRINHO/MADRINHA/BAPTISMO (godfather/
godmother/baptism).
AVÔ (grandfather)
FILHO (son)
FILHA (daughter)
(FEMININO+FILHO)
(feminine+son)
IRMÃO (brother)
SOBRINHO (nephew)
98
SIGN
COHESION
ETYMOLOCICAL PROPOSAL
SOBRINHA (niece)
(FEMININO+SOBRINHO)
(feminine+nephew)
Weak*
Old family ties between nephews/nieces and uncles/aunts. The latter were also
godfather/godmother to the first. These people were connected not only by a
consanguinity issue but also by the ties imposed by a religious ceremony: baptism.
Hence, the location and movement of SOBRINHO (nephew) is the same as
the location and movement in PADRINHO/MADRINHA/BAPTISMO (godfather/
godmother/baptism).
PADRINHO (godfather)
High-
Association with the movement from baptism – to poor water onto the baptized
person’s head.
MADRINHA (godmother)
(FEMININO+PADRINHO)
(feminine+goddfather)
High-
Association with the movement from baptism – to poor water onto the baptized
person’s head.
AFILHADO(godson)
(PADRINHO+FILHO)
(godfather+son)
Weak+
Comes from the signs PADRINHO/BAPTISMO (godfather/baptism), because the
godson is a “son” acquired through baptism.
AFILHADA (gaddaughter)
(PADRINHO
+FEMININO+FILHO)
(godfather+feminine+son)
Weak+
Comes from the signs PADRINHO/BAPTISMO (godfather/baptism), because the
godson is a “son” acquired through baptism.
SOGRO (father-in-law)
Weak-
Dactilology “I”, comes from the idea of “important in the family” (explination
given as a mere supposition, with no certainties).
SOGRA (mother-in-law)
(FEMININO+SOGRO)
(feminine+father-in-law)
Weak-
Dactilology “I”, comes from the idea of “important in the family” (explination
given as a mere supposition, with no certainties).
GENRO (son-in-law)
Null
----------------------------------------------------------------------------
NORA (daughter-in-law)
(FEMININO+GENRO)
(feminine+son-in-law)
Null
----------------------------------------------------------------------------
CUNHADO (brother-in-law)
Weak*
Association to the sign SEGUINTE/AO LADO (next/next to), which is manifested
by similarity in handshape to CUNHADO (brother-in-law).
CUNHADA (sister-in-law)
(FEMININO+CUNHADO)
(feminine+brother-in-law)
Weak*
Association to the sign SEGUINTE/AO LADO (next/next to), which is manifested
by similarity in handshape to CUNHADA (sister-in-law).
Medium*
Combinat6ion of the signs PAI+SEGUNDO (father+second), implying several
associations:
– a second father;
– a false father;
– a second person who gets married to the mother.
Medium*
Combination of the signs MÃE+SEGUNDO (mother+second), implying several
associations:
– a second mother;
– a false mother;
– a second person who gets married to the father.
Weak+
Combination of the signs FILHO+SEGUNDO (son+second), implying several
associations:
– a second son;
va false son, one who is not the true son.
ENTEADA (stepdaughter)
(FEMININO
+FILHO+SEGUNDO)
(feminine+son+second)
Weak+
Combination of the signs FILHO+SEGUNDO (son+second), implying several
associations:
– a second daughter;
– a false daughter, one who is not the true daughter.
NETO (gradson)
Weak-
Association to AVÔ e AVÓ (grandfather and grandmother), which are produced
with a chin location, for explaining this sign’s location at the chin.
NETA (granddaughter)
(FEMININO+NETO)
(feminine+grandson)
Weak-
Association to AVÔ e AVÓ (grandfather and grandmother), which are produced
with a chin location, for explaining this sign’s location at the chin
PADRASTO (stepfather)
(PAI+SEGUNDO)
(father+second)
MADRASTA (stepmother)
(MÃE+SEGUNDO)
(mother+second)
ENTEADO (stepson)
(FILHO+SEGUNDO)
(son+second)
Cognitive Studies on Portuguese Sign Language
(LGP): a work in progress
Estudos Cognitivos em Língua Gestual Portuguesa:
estudo de arte
Maria Vânia Silva Nunes1*, Paulo Vaz de Carvalho1, Ana Mineiro1 and Alexandre Castro Caldas1
1
Instituto de Ciências da Saúde da Universidade Católica Portuguesa
Abstract
Resumo
The main goal of present paper is to present the research
lines that we are presently following in the study of Portuguese
Sign Language (LGP), including the replication of previous studies
in other sign languages (such as a working memory study) but
also proposals that are, to our best knowledge, original (such
as the study of time perception in deaf). We start by making a
review of the theoretical framework that has supported several
studies in other sign languages, particularly in American Sign
Language (ASL), identifying then the need of replicating these
studies for LGP. The status of our works is also presented.
Keywords: Portuguese Sign Language (LGP), American Sign
Language (ASL), Working Memory, sign languages, cognitive
development, time perception, deafness. 
Deafness and cognitive development
There are no doubts that, although being the
result of genetic programming, the development
and maturation of the brain and related cognitive
development, are crucially dependent of the interaction with the environment. As a key feature of
this interaction are sensorial inputs, being now
well known that sensorial deprivation may result
in deficits in structural and functional organization.
Congenital sensorial deprivation, as the one that
results from congenital deafness or blindness, can
then, theoretically, have a deleterious effect in that
organization. Even when there are circumstances
where these deficits can be corrected later in life,
there is an increased probability of passing the
critical period without stimulation.
We can identify two different kinds of questions
when researching in deafness. The first concern
O objectivo principal deste artigo foi apresentar um estudo
que se encontra a decorrer, no ICS da UCP.
Neste trabalho, propomo-nos replicar estudos já feitos para
outras línguas gestuais, nomeadamente, relativamente à memória
de trabalho. As nossas propostas englobam também estudos
originais, nomeadamente no que concerne a percepção do
tempo nos surdos. Neste artigo fazemos uma breve revisão da
literatura relativamente ao Estudo da Arte nas outras línguas
gestuais, particularmente na ASL, identificando a necessidade
de replicação desses trabalhos para a LGP. O Estado do
desenvolvimento do nosso trabalho é também apresentado.
Palavra Chave: LGP – ASL – Mamória de Trabalho – Desenvolvimento cognitivo – Percepção do tempo pelos surdos
is to know how deprivation of auditory sensorial
inputs influences cortical and cognitive organization.
The second, behind the scope of this article, and of
greater relevance for intervention and rehabilitation,
concerns the existence of a critical period for the
auditory system. This second question, related with
the existence of a critical period for the auditory system, gains even greater importance when considering
that congenital deaf that are implanted as adults have
worst auditory performances and little gains in what
concerns their linguistic development. In fact, several
studies seem to suggest that auditory performance
of congenital deaf with cochlear implants is crucially
dependent on the age of implantation.
The importance of Sign Language
Despite sensorial deprivation, auditory deficiency
or deafness are multidimensional phenomena with
* [email protected]
[email protected]
100
several social, medical and educational features
needing consideration. Therefore children with
auditory impairment can follow various developmental pathways. Several factors can account for
this variability. When taught sign language as a
native language, deaf children are able to achieve
linguistic milestones at a proper age. That is probably
why one factor of good prognostic is when deaf
children have deaf parents, and are immersed in
that linguistic environment since birth.
However, sign languages and oral languages differ
in several relevant features namely in the sensorial
modality that it is used to convey the linguistic
information. In sing languages we have the linguistic
content that is understood as such by the brain,
which conveyed to the brain through a different
sensorial modality, that is a visuo-spatial modality
instead of an auditory one. That poses a problem to
the direct application of the classic working memory
model to sign language.
Working Memory Model
The relation between memory, particularly working
memory and oral language is well known, and has
been extensively explored in numerous studies.
In fact the working memory model postulates the
existence of a mechanism specialised in processing
phonological and linguistic information.
Working memory is an influential model that
was first proposed in 1974 by Baddeley and Hitch
in a seminal paper that points that “despite more
than a decade of research on the topic of short term
memory we still know virtually nothing about its role
in normal human information processing” (p. 48).
According to these authors the concept of short
term memory should be replaced by one that of a
working memory system. In this proposal we have
a control system with limited storing and processing
capacity, the central executive, and two subsidiary
systems, one used for verbal material, the other one
used for visuo-spatial material. In this initial paper
the authors suggest that working memory plays an
important role in several cognitive functions.
Phonological loop is the best studied component
of the model and it is related with the processing of
verbal and phonological material. Two components
were proposed as part of the Phonological loop:
a speech based store that keeps the mnesic trace
and an articulatory control process. It is assumed
that auditory information gains mandatory access to
the store. This loop explains several experimental
effects and its considerable explicative power, has
lead people to reflect about the functional role of
the phonological Loop in speech comprehension.
Subsequently it was proposed that the phonological
loop could play an important role in the long term
learning of unfamiliar material (Baddeley, Papagno
and Vallar, 1988). Its role in understanding the
process of language acquisition was also studied
(Gathercole and Baddeley, 1990), and relations with
the acquisition of skill and knowledge in reading
and mathematics are also found (Gathercole et
al, 2006). The phonological loop seems to be a
“Language Learning Device” to learn new words
(Baddeley, Gathercole and Papagno, 1998) that crucially relies on the phonological store. It also seems
to be related with a second language acquisition
(Service, 1992). Some studies show that phonological working memory capacity, as measured by
non word repetition, can influence the outcome of
cochlear implantation, even more than implantation
age (Willstedt-Svensson et al, 2004).
The proposal of a phonological loop to deal with
verbal and phonological material is substantially
validated. In fact, phonological loop explains several
experimental effects: phonological similarity that
shows that phonologically similar words are less
remembered than dissimilar items (Conrad and Hull,
1964), strengthening the idea of a phonological
coding. It also explains the irrelevant speech effect,
that shows that irrelevant auditory material disrupts
performance (Colle e Welsh, 1976) and reinforces the
idea that auditory material gains automatic access to
the phonological store, and the word length effect
according to which, lists of longer words are less
well remembered than lists of short words (Baddeley, Thomson and Buchanan, 1975), which is seen
as evidence for an articulatory rehearsal process
that seems to occur in real time. This word length
effect can be prevented by articulatory suppression
(Baddeley, Lewis and Vallar, 1984). Articulatory
suppression is obtained when subjects perform an
irrelevant mouth movement, while performing a
span task. Articulatory suppression also prevents
phonological similarity effect for visually presented
material since it prevents sub-vocal recoding (Murray,
1986). In the phonological loop people seem to
remember as many items as they can say in two
seconds (Baddeley et al. 1975), which reinforces the
idea that rehearsal occurs in real time.
The other subsidiary system, the visuo-spatial
sketchpad, is less studied. The visuo-spatial sketchpad
is seen as a store that, together with control proces-
Cognitive Studies on Portuguese Sign Language (LGP): a work in progress
ses, is responsible by registering the visuo-spatial
information and keeping it trough a rehearsal process.
More recently other component was added to the
model, namely the episodic buffer (Baddeley, 2000),
but the division between a verbal domain and a
visuo-spatial domain seems to reflect a fundamental
division in human cognition.
Working Memory Model and American
Sign Language (ASL)
Wilson and Emmorey (1997) studied the working
memory system of ASL signers considering the possible interference of the modality by which language
reaches the brain: visual or auditory. As pointed by
these authors many differences between verbal and
visuo-spatial working memory have been attributed
to differences between audition and vision. In fact,
audition is more related with time, and vision is
more related with space. As synthesized by the
authors there are strong unidirectional associations
between auditory items that are not found in visual
working memory that may demand a nonlinear
structure. That could imply that working memory
for sign language may differ systematically from
that for speech.
But the authors also stressed that there are several
similarities between sign and speech, namely that
signs are not holistic gestures but are constructed
from a set of meaningless units, combined in a
similar manner to that of the phonological level
in spoken languages (Battison, 1978 in Wilson &
Emmorey, 1997) and that there is a close relationship
between perception and production, unlike what
happens with the vast majority of visual stimuli that
are usually used to evaluate working memory, and
that allows rehearsal processes.
Based on these similarities, and in the growing
evidence at the time that ASL-based memory code
for temporary storage resembles the type of speech
coding used by hearing subjects, the authors went
further and directly explored the existence of a
“phonological” loop for sign language.
They found a phonological similarity effect (using
signs that were either similar or dissimilar in terms
of handshape), replicating previous results. They
also found a manual articulatory suppression effect
(obtained by asking the subjects to open and close
their fists, alternating the hands). They argue that
suppression and similarity seem to derive from
separate components of the system. For material that
needs recoding (pictures of easily named objects)
101
the similarity effect was eliminated by articulatory
suppression. They obtained similarity effect when
there was no hand motion, but hand motion suppressed the similarity effect, so it appears that an
articulatory process in needed to recoding. They
also found a “word” length effect; presenting short
signs and long signs, and showing that memory
performance is sensitive to the articulation time of
the signs. On the other hand, they also found that
suppression prevents the length effect. Synthesizing
their findings about the effects, and the interactions
between them, are analogous to what happens for
oral language, reinforcing the hypothesis of a “sign
loop”.
Working Memory Model and
Portguguese Sign Language (LGP):
A work in progress
The work of Wilson and Emmorey (1997) and posterior works seem to suggest that the “phonological
loop” is a mechanism that is related to the processing
of linguistic input, regardless of the modality of that
input, suggesting that working memory system is a
flexible one. Neuroscience data reinforces this similarity, for instance, recent neuroimaging studies had
found an inner sign analogous to the inner speech
and recent neurophysiologic evidence supports the
notion of an amodal site for carrying phonological,
syllable like representations in the temporal lobe that
it’s active when either speech or sign are processed
in syllable tasks (cf. Ronnberg, 2003).
Deaf brain organization, in a similar manner to
what happens in blind subjects, seems to display
cross modal plasticity (in the case of the deaf it
is when sign language activates areas of auditory
cortex). In fact is today known from imaging studies
that deaf subjects display auditory cortical activity
when they are processing sign language gestures
or other complex visual stimuli (Pettito et al 2000).
Multisensorial convergence seems then to be a generalized feature of brain and cognition (Bavelier et
al, 2006). Besides functional reorganization, auditory
experience and/or exposure to sign language during
human development seems to impact even anatomical organization. For instance volumetric analysis
with MR shows that deaf people present a greater
volume of grey matter in posterior left insular lobe.
(Allen et al. 2008).
These results seem to point to a universal response
of the brain and cognition to the learning of a
sign language, namely ASL. In the case of working
102
memory, we can think that this similarity of ASL with
spoken language reveals and underlying capacity
of the human brain and cognition to deal with an
alternative sensorial modality of linguistic input.
However, and more than in oral languages that
are well established, there are significant differences
between ASL and LGP.
ASL is the dominant language in the American
deaf community, in some parts of Canada (Anglican)
and in some regions of Mexico. The ASL is also used
in other countries like the Filipinas, Malaysia, Hong
Kong, Dominican Republic and Puerto Rico among
others. Just like other sign languages its grammar and
syntax is different from any other spoken language
within its area of influence. The actual ASL has
its origin in a confluence of various events and
circumstances and has four great influences: Home
signs; Martha’s Vineyard sign language; Indians tribe
sign language; and the Old French Sign Language.
Thus ASL has a grammatical structure based in these
four great influences.
Having completely different origins, the Portuguese
Sign Language (LGP) has a distinctive grammatical
structure from ASL. LGP is a language that is used
by Portuguese Deaf people in their communication,
being the most important window frame of their
identity. Introduced in Portugal by professor Per
Aron Borg, the great influence in LGP is Swedish
Sign Language. It’s possible that even before the
arrival of Professor Borg in Portugal, it exist an Old
Portuguese Sign Language developed in very little
communities spread around the country, but no fonts
confirm that hypothesis. Nowadays, the influence of
Swedish Sign Language in LGP can only be seen in
the hand configuration used in signs that describe
the Portuguese manual alphabet, very identical to the
Swedish one. Thus, other LGP linguistic parameters
are distinctive from the Swedish Sign Language,
probably as the result of the grouping of deaf children
and youngsters in residential schools.
Another great difference between these two languages is their historical educational background.
In 1880 in the Congress of Milan it was decided
by majority that deaf should be taught orally, that
sign language should be banished, and that deaf
teachers should be replaced by hearing teachers.
These recommendations had the opposition of two
countries: the EUA and Sweden that kept sign language as the method of teaching the deaf.
As consequence, for over a century, Portugal,
as several other countries, prohibited the use of
sign language in deaf schools, and LGP developed
outside school, namely in households, residential
households and deaf associations.
Therefore, while ASL was developing inside deaf
schools and in deaf communities, broadening its
lexicon and refining its grammatical structures, LGP
was only practiced in secrecy, becoming reduced to
a daily basis communication, being held back in its
developing. Although, in some periods in the USA,
the use of ASL in schools had lost some importance,
in the sixties, due to the linguistic investigations of
William Stokoe, ASL reacquired and enlarged its
statute as language for teaching the deaf students. In
Portugal, only in the nineties LGP became regularly
used as the language for educating deaf people
and only in 1997 was recognized by law as a true
language of deaf community.
Due to referred differences between the two
languages (ASL and LGP), their distinctive historical
background, and some other differences that seem
to exist namely in terms of phonological and lexical
variation.
As previously demonstrated younger sign languages (Aronoff et al., 2008, Israel and Sandler,
in press) appear to exhibit a greater amount of
sublexical variation (e.g. handshape). Differences
in social factors such as language age, size of the
Deaf Community and prescriptive language norms
seems to contribute to develop robust established
lexical and sublexical categories and signs with less
variation. Israel and Sandler study (in press) pointed
out meaningful differences in phonological variations
among categories in ASL, ISL (Israeli Sign Language)
and ABSL (Al-Sayyid Bedouin Sign Language). ASL
presented the less amount of variation, followed
by ISL and ABSL showed the most variation. This
study leads us to suggest that probably in LGP we
will find a greater amount of variation, than in ASL.
This is en important issue in order to design the
linguistic Stimuli in our work.
Some pilot studies conducted by Mineiro et al.
(2008) and Pereira et al. (in press) presented us some
evidence about the variation of polissemic forms and
the diachronic variation of forms in Portuguese sign
language lexicon. These findings were consistent
with the results of nominal criation in LPG found
by Henriques (2006).
Although these are preliminary studies summed
up to the fact that there is no established corpora for
LGP, our feeling is that they are concrete differences
in terms of lexical and sublexical variation between
LGP and ASL. Further research in this domain is
necessary to establish differences.
Cognitive Studies on Portuguese Sign Language (LGP): a work in progress
We consider that the studies made for ASL should
be replicated for LGP, and it is proposed that the
results of those studies may held some differences.
Presently, we are replicating the study of Wilson
and Emorey (1997) on (because there is no established corpora for LGP there are few concrete data
in several linguistic parameters) it was considered
that the studies made for ASL should be replicated
for LGP, and it is proposed that the results of those
studies may held some differences.
We are presently replicating the study of Wilson
and Emmorey (1997) on Working Memory because,
although being a relatively old study, it relies on
a well established model and holds important
conclusions regarding the nature of the cognitive
organization for sign language.
The materials are already constructed and implied
the arrangement of several lists of similar items,
dissimilar items, short signs, long signs and easily
nameable pictures. Building these materials has
proven to be a particularly difficult task. We have also
introduced a digit span task. Because a consistent
finding in working memory for sign language is that
storage capacity is significantly lower for sign that
for speech (Emmorey and Wilson, 2004), because
Portuguese signs with all the required characteristics to build the list (and with additional semantic
constraints) were hard to find, and based on the
opinion of experienced deaf teachers, our lists had
a maximum span of six items. However in some
tasks, when we did the pre-validation of our material
with elements of the deaf community we found, a
ceiling effect.
We can speculate that this ceiling effect was found
because probably there are differences in the relation
between the articulation length of ASL and LGP and
their correspondent oral languages that impact span,
or in other linguistic features.
We are now increasing the maximum number of
items in our lists in order to proceed to the applications in the experimental group. The experimental
group, including only native speakers of LGP, is
identified, and has already signed consent forms.
Experimental results are due until the end of the
current year.
Time Perception and Portuguese Sign
Language (LGP): A relation to explore
It is interesting to notice that the span size difference between sign and speech that occurs in immediate serial recall doesn’t occur for more complex
103
tasks (Boutla et al., 2004 in Emmorey and Wilson,
2004). A possible explanation that is put forward
is that sign languages rely in visual processing that
occurs simultaneously and don’t rely in temporally
encoded distinctions (cf. Emmorey and Wilson, 2004).
The different pattern that hearing and deaf subjects
present in forward and backward digit span, with
the backward digit span being more deleterious
to hearing subjects is interpreted as reflecting the
specialization of the loop in for the exact repetition
of a sequence of items in a given order (Wilson and
Emmorey, 1997) constituting one aspect in which the
authors identified differences between the speech
and the sign loop.
Taking into consideration these results, and beyond
the replication studies in course for LGP, we are
also exploring the hypothesis that these differences
in the reliance in temporal sequences may have an
impact in time perception.
Time perception and its measuring are basic
components of cerebral function, and temporal
processing is an integral part of many everyday goal
oriented behaviours. The explanation for subjective
time also assumes the existence of an internal-clock
mechanism. Previous work has shown an association
between the estimation of time and short term
memory/ working memory (Coelho et al, 2004). We
are presently adapting the tasks and procedures used
in this study to LGP in order to address the relation
between sign language (LGP), time perception and
working memory.
Recognized as the official language of the Portuguese deaf community only in 1997, LGP is a
language that presents a wide research field to which
we expect to contribute in a productive manner.
References
1.
Allen J., Emmorey K, Bruss J and Damasio H. (2008) Morphology of the
Insula in Relation to Hearing Status and Sign Language Experience. Journal
of Neuroscience, November 12, 2008, 28(46).
2.
Aronoff, M., Meir, I., Padden, C. A., & Sandler, W. (2008). The Roots of
Linguistic Organization in a New Language. Interaction Studies, 9, 133-153
3.
Baddeley, A.D. & Hitch GJ (1974). Working Memory. The psychology of
learning and motivation (Bower, G.A. Eds). Academic Press. New York
(vol 8, 47-89).
4.
Baddeley, A.D., Papagno, C. & Vallar G. (1988) When long term learning
depends on short term storage. Journal of Memory and Language, 27,586595
5.
Baddeley, A.,Thomson . N, Buchanan, M. (1975) Word length and the
structure of short term memory, Journal of Verbal Learning and Verbal
Behavior, 14, 575-589
104
6.
Baddeley, A. D., Gathercole, S. E., & Papagno, C. (1998). The phonological
loop as a language learning device. Psychological Review, 105, 158-173
7.
Baddeley, A.D. Lewis V. ,Vallar G., Exploring the articulatory loop. Quarterly
Journal of Experimental Psychology 36A (1984), 233–252.
8.
Baddeley, A.D. (2000) The episodic buffer: A new component of working
memory? Trends in Cognitive Sciences. 4, 11, 417-423.
9.
Bavelier, D., Dye, M.W.G., & Hauser, P. (2006). Do deaf individuals see better?
Trends in Cognitive Sciences, 10, 512-518.
10. Coelho M. , Ferreira J.J , Dias B. , Sampaio C. , Pavão Martins I. , CastroCaldas A. (2004) Assessment Of Time Perception: The Effect of Aging;
Journal Of The International Neuropsychological Society (2004), 10, 332-341
11. Colle, H. A., & Welsh, A. (1976). Acoustic masking in primary memory.
Journal of Verbal Learning and Verbal Behavior, 15, 17-32
12. Conrad, R. & Hull, A.J. (1964) Information, acoustic confusion and Memory
Span. British Journal of Psychology, 55, 429-432
13. Emmorey K., Wilson M. (2004) The Puzzle of working memory for sign
language. Trends in Cognitive Sciences, 8,12, 521-523
14. Gathercole, S.E., Alloway, T.P., Willis, C.S., & Adams, A.M. (2006). Working
memory in children with reading disabilities. Journal of Experimental Child
Psychology, 93, 265-281.
15. Gathercole, S. E. & Baddeley, A. D. (1990). The role of phonological memory
in vocabulary acquisition: A study of young children learning new names.
British Journal of Psychology, 81, 439-454
16. Henriques, L. (2006) Para uma descrição da variação nominal em LGP,
MA Diss. Lisbon (n/p.)
17. Israel, A, Sandler, W. (In Press) Phonological category resolution: a study
of handshapes in younger and older sign languages. Cadernos de Saúde,
Edição Especial.
18. Mineiro, A. Duarte, L. Carvalho, P. Tebé, C. e Correia, M. (2008)Aspectos da
Polissemia nominal em Língua Gestual Portuguesa. Polissema, 8, 37-56 .
19. Murray D. (1968) Artculation and acoustic confusability in short term
memory. Journal of experimental psychology, 78, 679, 684
20. Pereira,J. Morais, I. Duarte, L., Mineiro, A. (In Press) Diachronic Variation in
portuguese Sign Language. Proceedings of the 1st Symposium in Applied
Sign Linguistics, University of Bristol, Centre for Deaf Studies, Bristol: UK.
21. Petitto LA, Zatorre R J., Gauna K,. Nikelski E. J, Dostie D, and. Evans A C
(2000) Speech-like cerebral activity in profoundly deaf people processing
signed languages: Implications for the neural basis of human language.
PNAS 97:13961-13966
22. Ronneberg, J. (2004) Working memory, neuroscience and language –
evidence from Deaf and hard-of-hearing individuals, (eds.) Marschak,
M. Spencer, E. (2003) Oxford Handbook of Deaf Studies, Oxford University
Press, chp34: 479-489.
23. Service E. (1992) Phonology, working memory, and foreign-language
learning. Q J Exp Psychol . Jul;45(1):21-50
24. Willstedt-Svensson U; Löfqvist A; Almqvist B.; Sahlén B (2004) Is age at
implant the only factor that counts? The influence of working memory on
lexical and grammatical development in children with cochlear implants
International Journal of Audiology: 43 (9) 506 – 515
25. Wilson M., Emmorey K. (1997) Working Memory for sign language: A
window into the architecture of the working memory system, Journal
of Deaf Studies and Deaf Education 2:3, 121-130

ESPECIAL Línguas Gestuais

Transcription

Similar documents

Time: 6:00 to 8:00 PM Where: Shakey`s Pizza Parlor 12540 Valley

Pedro Fernández began his international career when he was

Dedicado a todas las personas que creen en la musica

August newsletter for webpage - Tyler Deaf and Hard of Hearing

THE HAWK - english

el libro de colorear, Jaime, el Dulce Guerrero

TODAY`S READINGS First Reading — When the wicked turn away

January 17, 2016

July 03, 2011 - St. Henry Catholic Church

093. Matías Ch`ok cuenta que daños causó un brujo a su familia