O n f e r e n c e

Short paper 
Background 
One of the primary goals in L2 learning and teaching is to improve the comprehension of L2 
texts and discourse. To achieve successful comprehension, learners are taught various 
aspects of the target language. Among them, vocabulary has been consistently reported to 
play a key role in achieving L2 proficiency (Ellis 1995; Hatch and Brown 1995; Hu and Nation 
2000; Lewis 1993; Schmitt 2008). Therefore, since the 1990s, increasing attention has been 
paid to how L2 vocabulary is most effectively learned (e.g. Chen 2009; Laufer and Hulstijn 
2001) although the best way has not yet been found (Schmitt 2008).
Amid the alternatives, the use of pictorial aids in a CALL environment could be advantageous 
in explaining the meanings of L2 words in dictionaries or learning materials in annotations or 
glosses (Lomicka 1998). In particular, the use of pictures as visual glosses has been shown 
to facilitate longer retention of target L2 vocabulary (e.g., Chun and Plass 1996; Al-Seghayer 
2001; Yoshii and Flaitz 2002; Yeh and Wang 2003). Our study, however, questions the 
generalization that visual glosses are always effective in L2 vocabulary learning by 
considering the three main concerns: vocabulary items, configurations of the visual glosses 
(Sato & Suzuki, 2010, 2011, 2012) and individual factors. 
Our Concerns
The first concern is the type of vocabulary item that would be best suited for visual glosses. 
Previous studies might generalize the advantage of visual glosses in L2 vocabulary learning 
withiout much consideration of kinds of vocabulary items. This might be because they 
conducted their research in terms of incidental learning. That is, reading texts were initially 
conducted, and then vocabulary tests were employed to examine how much vocabulary was 
incidentally retained. Successful vocabulary learning for these studies means how much and 
how long the target vocabulary can be memorized incidentally in the reading text. As a 
result, studies of this kind might not pay attention to the differences among vocabulary 
items, e.g., semantic structure. Some vocabulary items are easier to learn becauase their 
semantic structure is relaviely simpler in the sense that each of their forms contains only one 
sense. However, if a target word has several sense in one form, learners might encounter 
difficulty learning all the senses and selecting the appropriate one according to the context. 
Recent studies (Sato & Suzuki 2010, 2011, 2012) confirmed the above observation and even 
demonstrated a less positive effect of visual glosses on the learning of L2 English prepositions 
in text and discourse comprehension and production. We thus wonder whether pictorial aids 
are always effective in any setting and what types of vocabulary items better accomodate the 
use of visual glosses.
Our second concern is related to the type of visual glosses, i.e., the configurations. Al-
seghayer (2001) shows that animation is a more effective technique than the use of still 
images. Littlemore (2009) also points out that three-dimensional diagrams are useful when 
they could be displayed dynamically. Therefore it might be thought that the more 
technologically-enhanced visual glosses learners are exposed to, the better learning results 
they could get. Sato & Suzuki (2010) examine the relative effects of two types of visual 
glosses, still images and animated images, in learning English prepositions. The figures below 
are the visual glosses of the preposition ‘over’. The left image is a still picture which 
conceptually depicts the prototypical sense 
based on Lakoff (1987)’s above
-across schema. 
The right image, on the other hand, depicts the same above-across schema but in an 
animation format. After showing these glosses, a vocabulary test was conducted to examine 
which type of visual aid best develo
ps the understanding of the senses of ‘over’. 

---

-301- 
2014 CALL Conference 
LINGUAPOLIS
www.antwerpcall.be 
Visual glosses of ‘over’ (Sato and Suzuki 2010)
The findings conclude that no significant difference between the two glosses is found, 
although significantly different learning results are found between visual glosses and verbal 
glosses (which verbally explains the protopypical sense above). Sato and Suzuki (2010) 
indicate that the effectiveness of visual glosses for L2 vocabulary learning may also 
correlate with individual facotrs. 
The final concern is the individual factors in effective use of visual glosses in L2 vocabulary 
learning. We are wondering whether every learner could get a positive effect with 
multimodal tools as previous studies stressed. Some learners might easily and quickly 
visualize an image of the target word with the help of visual glosses, while others might not. 
For example, some learners might quickly visualize “The airplane is flying 
over
the 
mountain”, “My house is 
over
the river”, or “You will get over a problem” with the help 
of 
the glosses above, whereas others may still find it difficult to connect the the target word in 
the sentences with the images. The ease of visualization leads us to focus on the difference 
in learners’ cognitive styles. Some learners, for example, might
be better at conceptualizing 
knowledge with the help of visual glosses, whereas others might be good at analyzing 
knowledge through verbal information.
These inclinations can be captured by grouping the cognitive learning styles as holistic 
cognitive st
yle or analytic cognitive style (Littlemore 2001), or the learners as “imagers” or 
“verbalizers” (Riding and Rayner 1998). Littlemore (2001) shows that those who have a 
holistic cognitive style, or better metaphoric competence, could derive metaphorical senses 
faster than those who have an analytic cognitive style. Grounded on the above, we further 
investigate the correlation between learners’ cognitive style and the effectiveness of visual 
glosses in learning L2 vocaublary.
Besides the cognitve style, t
he differnece in learners’ first language (L1) is another 
individual factor to be examined. English prepositions are regarded as difficult to learn 
because L2 learners, despite being advanced learners, do not always understand their 
meanings (Brala, 2002; Lindstromberg, 2001). This might be especially true for Japanese 
learners because translations which are added to each sense of the target word may 
prevent learners from understanding the semantic networks of the senses.

---

-302- 
2014 CALL Conference 
LINGUAPOLIS
www.antwerpcall.be 
For example, a protypical sense of 
over like “an airplane is flying 
over
the mountain” is 
translated into “hikouki ga yama 
no ue wo
tondeiru”, but the translation “no ue wo” is also 
used in the protypical sense of “above” and “on”. This overlapping might lead to appropriate 
use of the senses according to the context. However, we are wondering whether this 
difficulty would be applicable to learners whose L1 is not Japanese. For example, Chinese 
learners of English, whose L1 linguistic features are very different from that of Japanese but 
are rather similar to English, might not experience the same difficulty.