Data can be difficult to interpret into meaningful conclusions however this study explores several questions which can help direct future research in this area. For example, the
data let us know that both native and naïve signers do cluster signs into category-like areas and to different extents. They however do not let us know a definitive answer to
what those categories are. We can see that even without any knowledge of ASL, naïve participants were able to use the primes to make guesses as to which categories may exist,
however to what extent is not ascertainable from this study. We can see by the variability of the slopes in the average charts (Appendix D) which categories may be preferred by
native signers for future work. However, looking at the level of the first category in the Street condition for native signers may be misleading as well. The majority of
participants selected either the first or third token consistently, not some combination of both, whereas both groups selected the 80º token as the appropriate token for the
concept of 'forward'. This is alluded to by the response angle histograms in Fig. 5 (Appendix E).
Results of the identification paradigm seem likely that in the two conditions a different token is preferred for the '12 o'clock' orientation and the '
forward' orientation, thus it would be ideal to continue to allow the participants to select this. Also, if we do not account for individual dialect differences, different
category choices (like those explained above in the Street condition), would not be detected and what is a border to one individual would not be easily discernable from a
category for another. This of course can muddle the results and increase likelihood of a Type II error. One way to account for this might be to ask the participants to select
the specific categories from the prime before asking them to participate in the task. Once selected, the participants will only have their personally selected categories to
choose from (as in other ASL CP studies) and will make identification of category and border ranges much clearer and more definitive.
One final major weakness that was not accounted for in the identification task portion of this study, is the likelihood that the participants were trying to match the signer
's palm orientation exactly rather than choosing a sign that matches the intended meaning as they would most naturally receive (and produce) it. Though it was hoped that
the participants would use their knowledge of ASL to select a prototypical exemplar for each category, there is no way of knowing with these singular primes if that is the case.
One way to counteract that weakness could be to have the participants look at multiple primes of the same story signed by different signers. Done that way before the selection
of category labels, the phonemic entry for each sign would be more likely activated for the participants rather than the allophonic production by the individual signers in the
priming material. If successful, the participants would likely choose the phoneme exemplar they feel best represents the sign, thus a more accurate category label could be used
in the research.
To the question of how discrimination varies along the continuum, again no conclusion can be made. As the response acuity varied for all conditions in the exact same manners, it
may indicate a variable perceptual space along this continuum, but it is more likely influenced by other factors, primarily stimuli weakness. Though the utmost care was taken to
provide video stimuli in correct and uniform angles, there were other non-linguistic visual cues that participants could use to discriminate between tokens, miniscule changes in
lighting, sign rate, hand positioning, clothing arrangement, etc. can indicate to participants, consciously or otherwise, which of the videos is not like the test stimulus. One
way to overcome this problem would be to use multiple tokens of multiple signers at each angle. This would cause all video pairings to have miniscule differences and for the
participant to use the parameter in question as the decision criterion.
Approaching the final question of contextual influence on discrimination, we do find support for context primes influencing discrimination at a single point. Though the stimuli
is weak and likely a confounding factor in making other types of conclusions, at the 30º - 40º range, there was a significant difference in discrimination for Native signers
between the Street and Clock contexts. Because of the type of data acquired by the identification task, it is unreasonable to make any conclusions about the categorical nature
of this difference; however the same stimuli were used for both conditions and only the prime was manipulated. It is not unreasonable to find support for contextual influence on
the discriminability of this continuum.
As stated above, one of the most productive ways to improve this study would be to use multiple signing models and multiple tokens from each to better increase construct validity.
However, these methods of stimuli creation require significant time, resources, and effort by a team of dedicated individuals. This is not to say that it is unlikely in the
proper context, however the exploratory nature of this study and scarcity of resources did not allow for these changes to be made during execution of the study.
ASL uses the affordances of a three-dimensional articulatory space and multiple external articulators to its most effective advantage. The contexts provided in this study address
not only a difference in predicted categories along a continuum but also different uses of space which cannot be represented as such in linear spoken language. The Street context
made use of what is referred to as 'neutral' space by linguists of ASL (Liddel, 2003). This type of use does not make reference to the space around the signer, nor is
it contrastive in undefined increments. There are a few ways that an individual may sign 'forward' and 'left' however, if a signer signs 'left' with
the orientation a little more forward, or a little more vertical, the meaning does not change, in theory or in practice.
In the Clock condition, the signs used a type of space called 'depicting space'. One characteristic of depicting space is that it allows the signer to depict the
arrangement of items within an established space and theoretically can be considered contrastive to an indefinite level of accuracy. The signs from the continuum in the Clock
scenario were created in this type of space. If they were produced slightly differently, it is possible that the audience may interpret the meaning differently. For example, if
the prime had signed the hand position at exactly 45º, between 10 o'clock and 11 o'clock positions, it may be interpreted that the hand itself was in reality positioned
between those two locations.
One open question in perception of parameters of ASL which has not been—and truly may forever not be—resolved is which cues are used to identify each. One speculative hypothesis
that one may posit about perception of palm orientation is that when it occurs in a depicting space it is perceived in the same absolute manner and governed by the same
mechanisms that govern location contrasts, or perhaps even non-linguistic visual images. The other side of that hypothesis would be that when perceived using neutral space as a
lexical sign, it is perceived with the same mechanisms as handshape and therefore is influenced by linguistic experience.
Another hypothesis about the contextual difference could be that the Clock prime took advantages of palm orientations that are not often used. If that's the case, then it is
intuitive that palm orientations used as frequently as the cardinal directions used in the Street prime would have a much stronger influence on linguistic perceptual ability than
those used with low frequency by native signers.
As understanding of speech perception advances, it is important that we understand which elements are universal and which elements are unique to the linguistic pathways. By
comparing aural-oral pathways with visual-signed pathways we can much more thoroughly understand the phenomena which we study. Furthering research in work like this will much
more effectively allow understanding of the scope of language.
