4.3 Distribution of Multi-word Hedges across Three Corpora
4.3.2 Anticipatory it Clauses
4.3.2.3 It + adjective + that / It + adjective + to + infinitive + that
Comparing the total number of use across three groups, generally speaking, L2 writers appear to use less number of adjectives to present their claims with only 4.98 per hundred thousand words. On the contrary, both native and journal writers would employ greater number of hedges with 16.41 per hundred thousand words for natives and 14.11 per hundred thousand words for journal writers. Native and journal writers show great similarity in the overall frequency of adjectives.
Of all the adjectives shown in the Table 4.9, the highest number of occurrences is possible, which consistently ranks first across three corpora, but it is obvious that L2 writers have lower frequency in comparison with the other two groups. Both native and journal writers share quite the same result with 5.51 per hundred thousand words for native students and 6.69 per hundred thousand words for journal writers, but L2 writers only obtained 1.82 per hundred thousand words, which is relatively low in frequency compared with the other two groups. Other observed differences on the three groups of people are also shown on the words such as clear and likely. L2 writers also display a limited range of adjectives, whereas native writers and journal writers have a wider range of vocabulary use. Native and journal writers would use
70
other instances such as likely and conceivable. In addition, they also employ other syntactic constructions, for example, it is reasonable to assume that… it is safe to say that… it is safe to assume that… to attenuate the strength of the proposition. This strategy is presented in the form of an adjective followed by an infinitive clause, which is hardly observed in L2 corpus. But this tendency still requires more exploration since the frequency is not high. Overall, as far as the data concerned, their use in terms of structure seems to be more diverse than L2 writers. This evidence might reflect that native and journal writers have greater storage of adjectives. A plausible explanation to L2 writers’ low frequency of the adjectives could be the result of less exposure to these devices. This infrequent use of adjectives might possibly due to the fact that the learning materials did not put much emphasis on the promotion of other potential adjectives or adjectival syntactic constructions which could function as hedging devices. The following are some of the example sentences derived from the three corpora.
(i) Fourth, according to the data of interviews, many participants attributed their decrease of motivation to teachers. It is possible that students did not fairly report teachers' behavior due to their biases against certain teachers. (Providence 04) –L2 corpus
(ii) It is likely that some of the participants might not feel interested in reading some of the assigned novels. (Tunghai 04) –L2 corpus
(iii) I knew that I preferred to keep the group small (ten participants or less) so that each participant would have a voice and have time to share. With too many participants in book discussion setting, it is possible that some students would not speak out. So, in the spring of this school year, I chose four students to participate in my study. (Ohio State 40) –Native corpus
(iv) It is possible that this participant was already aware of his issues with attention and the training had no effect on his level of awareness. However, it is
71
equally likely that the same learner would not have been aware of the effects of wandering attention on listening comprehension prior to the training. (Ohio 36) – Native corpus
(v) As both groups used the course textbook it is possible that the experimental group may have simply studied more for the unit tests than the control group, which could have affected their scores and hence the later statistical analysis based on those scores. (CALL 12) –Journal corpus
(vi) With less glossing, it is likely that some readers will not take the time to decipher the meaning of unknown words, especially if the storyline is understandable without the effort. (FLA 93)–Journal corpus
Table 4.9 Frequency of Adjectives in Anticipatory it Clauses
L2 RF NF Native RF NF Journal RF NF
Note: RF=Raw Frequency NF=Normalized Frequency per 100,000 words
72 4.3.3 Abstract Entity
Another widely used hedging strategy is abstract entity. As suggested earlier in Chapter Two, any inanimate sources can be classified into this category (Hyland &
Tse, 2005). The present study has discovered some of the inanimate sources such as research findings, results, and studies. These inanimate subjects are found more frequently in active rather than in passive forms, which is in similarity with Master (2002) who analyzed 2979 pairs of inanimate subjects with animate verbs in scientific articles. The predominant use in the active form of abstract entity + main verb is not surprising to find as it has been considered that this linguistic feature is a quite typical marker in academic writing (e.g. Master, 2002).
Inanimate sources rarely co-occur with sensory verbs (e.g. seem, appear).
Cognitive verbs and non-factive reporting verbs occur more frequently. The following section thus will include only a discussion on the use of the two major verb categories.
a. Cognitive Verbs
Table 4.10 displays that in L2 corpus the verb imply is 4.62 per hundred thousand (80%, 4.62 / 5.35) which comprises more than half of the total number of cognitive category. This means that L2 writers seem to heavily rely on this verb to make implications of the research findings. Taking the result to compare with the results from the other two corpora, the data reveals the quantity of cognitive verbs collocated with abstract entities is relatively small, whereas L2 writers seem to be at a slightly higher frequency. Cognitive verbs mainly serve to express one’s personal beliefs, attitudes, which is not surprising that it is rare to find cognitive verbs collocated with abstract entities. A number of studies (e.g. Hyland, 1998a, 1998b) have found that
73
abstract entities tend to co-occur with verbs mostly related to research activities such as findings and results. This discrepancy in the frequency is interesting in the way that it tells us L2 learners might not have enough knowledge with the collocation of certain verb types and subjects.
b. Non-factive Reporting Verbs
Non-factive reporting verbs serve to report one’s research findings according to Varttala’s definition (1999). Three groups of writers exhibit quite similar use of non-factive reporting verbs. As can be seen from below, Table 4.10 displays some overlap in the use of these verbs. Indicate, show, and suggest rank the top three verbs in three corpora. In the area of non-factive reporting verbs, the current results are shown as in a number of studies (e.g. Hyland, 1998; Charles, 2006). In this study, lexical verbs (e.g. suggest, indicate, estimate) are frequently found in combination with inanimate subjects (e.g. findings, result). Research nouns such as result(s), finding(s), study tend to co-occur with non-factive reporting verbs (e.g. indicate, suggest, show). Below present some examples extracted from the corpora.
(i) The results indicate that learners who held these five learning styles could significantly apply metacognitive strategy. (NCUE 01) –L2 corpus
(ii) This result suggests that the teacher made every endeavor to model the functions the students did not use, hoping that the students may increase the ratio of these function usage. (Chung Cheng 03) –L2 corpus
(iii) The qualitative results indicate that, despite no statistical difference in comprehension, the iPad is highly motivating to children. Survey results revealed that…(BGSU 11) -Native corpus
(iv) Results from the questionnaires suggest that students understood the purpose of research, a broad concept to conducting it…(Dayton 24) -Native corpus
74
(v) For ungrammatical items, accuracy was 48.8% and was not statistically above chance…These results indicate that learners’ performance on this task was driven by accuracy on grammatical items…(SSLA 82) – Journal corpus
(vi) The results in the current study suggest that EGAP writing instruction can lead to transfer and such transfer… (JSLW 23) – Journal corpus
The structure of inanimate subjects with cognitive verbs is common in academic writing just as common as the anticipatory it clauses. From the examples above, this structure is frequently used in reporting researchers’ results, graphic findings, suggestions for future actions, and pedagogical implications. This structure enables writers to express tentativeness by attributing responsibility to another entity and thus reduce one’s personal commitment to the proposition. It gives readers a sense that that the “knowledge is constructed on the basis of external evidence” (Charles, 2006:501).
Three groups of writers consistently favor the use of research nouns such as the (present) study, the finding(s), the result (s), the research, the data to obscure their self-identity, which can be seen as a way of protecting themselves or lowering their personal responsibility. This finding is also supported by some of the previous studies (e.g. Hyland, 1998b; Charles, 2006) which as well discovered that-complementation clause helps reduce a writer’s presence in a text.
75
Table 4.10 Frequency of Verbs in Abstract Entity Clause
L2 Native Journal
Cognitive RF NF Cognitive RF NF Cognitive RF NF
imply 38 4.62 predict 8 0.98 predict 15 1.79
assume 6 0.73 imply 3 0.37 assume 4 0.48
assume 1 0.12 imply 7 0.84
posit 1 0.12
add 1 0.12
hypothesize 1 0.12 Subtotal 44 5.35 Subtotal 12 1.47 Subtotal 29 1.80 Non-factive RF NF Non-factive RF NF Non-factive RF NF
indicate 183 22.23 show 68 8.33 suggest 130 15.55
show 131 15.91 suggest 56 6.86 show 106 12.68
reveal 73 8.87 indicate 46 5.64 indicate 64 7.70
suggest 68 8.26 reveal 30 3.68 reveal 36 4.30
demonstrate 36 4.37 demonstrate 28 3.43 found 25 3.00
found 21 2.55 found 15 1.84 demonstrate 24 2.87
state 8 0.97 state 10 1.23 state 19 2.27
conclude 6 0.73 conclude 8 1.00 report 6 0.72
assume 6 0.73 report 4 0.49 conclude 2 0.24
report 1 0.12 note 3 0.37 recommend 1 0.12
observe 1 0.12 propose 1 0.12
recommend 1 0.12
Subtotal 534 64.86 Subtotal 270 33.11 Subtotal 413 49.45 Total 578 70.21 Total 282 34.58 Total 442 51.25
76 4.3.4 Human Subject
Figure 4.2 Frequency of Structural Patterns in Human Subject Clauses
In the earlier sections 4.3.2 and 4.3.3, the clauses which begin with anticipatory it and abstract entity have been discussed. Writers choose them to conceal their identity and to make statements or ideas more objective in order to be accepted by the academic community. Another hedging strategy is human subject clauses. Human subject clauses in this study fall into three sub-patterns, which are determined by the type of subject. They are first-person clauses, nominalized subject clauses, and named researcher clauses. Of three clauses, the clause type that begins with first-person pronouns is the least frequent, but native and journal writers showed slightly higher number of instances. For nominalized subjects and named researcher subjects, both native and journal writers are closer in terms of frequency. The overall results suggest that clauses which begin with first-person pronouns might not be popular in academic writing especially for L2 writers. People across three groups still prefer to use clauses which begin with nominalized subjects and named researcher to show objectivity and detachment.
77 4.3.4.1 First-Person Pronoun (I/We)
The use of first-person pronouns (I / we) does not show consistency in three groups. First-person pronouns in L2 corpus appear in low frequency with merely 2.18 hundred thousand words. The frequency in native corpus is higher with 4.53 per hundred thousand words. Journal writers achieved the highest frequency with 6.00 per hundred thousand words. In other words, the results reveal that L2 writers have the least number of this type of hedge and native writers come somewhere in between the L2 and journal writers.
In native corpus, the singular form I is especially popular, which might indicate that writers are more confident with their claim and that they are more willing to project their voice into writing. The singular I is generally associated with cognitive verbs such as believe. This finding is also in harmony with (Hyland, 1998b; Varttala, 2001) who also believe that the first person subject tends to go together with verbs of thinking or cognition.
(i) From the data collected in my research, I believe that the purpose of the composition classroom is to make students better writers by making them better communicators…(Kent State 28) –Native corpus
(ii) I believe that the L2 students’ positive attitudes towards the way content is covered in a LRS-based class may be directly related to how they perceive the communicative language classroom implemented at the school (CALL 18) –Journal corpus
Comparatively speaking, native writers also use inclusive pronoun we to engage readers into the text in order for them to feel that they are part of the text and that they share the same knowledge with the writer. Prior research studies also agreed that the use of inclusive we could help establish a close relation between reader and writer (e.g.
78
Hyland, 1998a). This hedging device is often used when the writer is presenting the evaluations of their own results. By establishing common background knowledge with readers, it can be considered as an act of politeness. Readers are being treated as intelligent beings who are allowed to have the freedom to form their opinions without the need to fully accept the arguments proposed by writers. In the following examples, it is evident that they use the construction we (can) see that to indicate assumed common knowledge in presenting or evaluating results.
(i) From these results, we can see that composition instructors do not form their courses to the exact specifications outlined by the university. (Kent State 28) – Native corpus
(ii) …we can see that Fred clearly responded to his students, but it is interesting to note that in table 3 that his comment dwindles with each passing student…(Dayton 22) - Native corpus
L2 writer corpus, however, is slightly different from the other two corpora. The use of first person pronouns (I, we) is found to be quite small in quantity. Most of the writers tend to use first person plural we instead of singular form I. The evidence might possibly suggest that L2 writers are less willing to show the subjective tone in the form of I. One possible explanation might be that the use of I has the sense of ownership (Harwood, 2005). They are afraid that their tone of voice would sound too strong that they would be considered as unacceptable or intrusive in the professional community. For L2 learners, the use of we instead of I could serve as a protective device, making a writer’s identity less visible and thus reduce a writer’s responsibility for the proposition.
In terms of verb use in we + verb + that, only a few verbs in cognitive category are discovered. L2 writers prefer to select the cognitive verb know which is the most
79
frequently used verb in this type of clause. It only occupies a small proportion with 0.61 per hundred thousand words which is relatively small in quantity compared with the other two corpora.
(i) As teachers, we all know that the more vocabulary learners are capable of remembering, the better reading performance they have. (NTUE 02) -L2 corpus
(ii) As teachers we know that there are a lot of ways nowadays to evaluate the learning of our students. (Chung Cheng 04) -L2 corpus
Table 4.11 Frequency of Verbs in First-Person Clauses
L2 Native Journal
Cognitive RF NF Cognitive RF NF Cognitive RF NF
know 5 0.61 believe 13 1.59 believe 12 1.44
see 5 0.61 expect 8 0.98 assume 11 1.32
imply 2 0.24 see 6 0.73 expect 9 1.08
conclude 2 0.24 predict 5 0.61 predict 6 0.72
infer 2 0.24 consider 2 0.24 hypothesize 5 0.60
expect 1 0.12 think 2 0.24 hope 2 0.24
hope 1 0.12 hope 3 0.37 see 5 0.60
Subtotal 18 2.18 Subtotal 37 4.53 Subtotal 51 6.00
Total 18 2.18 Total 37 4.53 Total 51 6.00
80 4.3.4.2 Nominalized Subject
Unlike first-person clauses, nominalized subjects help a writer remain implicit and stay behind the statement. The purpose of the structure is in great similarity with that of anticipatory it clauses. In this study, various nominalized subjects are found such as author, researcher, teacher. A nominalized subject usually refers to a third person, but sometimes it might refer to the writer himself or herself. Thus, it could be found either in literature or results section.
a. Cognitive Verbs
Generally speaking, the total number of cognitive verbs in this category varies from each group. The highest group is L2 writers followed by native and journal writers. In L2 corpus, people tend to employ believe and thought to convey people’s beliefs or opinions in general. From Table 4.12 as shown below, it is clear that these two verbs (4.85 per hundred thousand words) occupy over half of the total number of category (70%, 4.85/ 6.55). In native corpus, it shares almost the same result as L2 writers except that the verb thought is absent, which might be justified by the reason that thought is a spoken lexical verb which would appear more frequently in spoken than in written genre. (Holmes,1984). Thus, it could be concluded that believe is a verb that is shared across three groups, and thought was only apparent in L2 corpus.
b. Sensory Verbs
For sensory verbs, three groups of writers have shown quite similar results. As far as the data concerned, only one sensory verb feel is identified in this structure. The verb is largely used to report the participants’ views or attitudes toward experimental procedures, teaching and learning processes instead of researchers’ own point of view
81
or belief. In other words, the responsibility of the statement is attributed to a third person not including the author himself or herself. Thus, the verb tends to co-occur with subjects referring to third-person singular and plural nouns such as students, participants, teachers and learners. Below are some of the examples from the corpora.
(i) …the great majority of proportions of the participants felt that they had more positive attitudes toward reading after participating extensive reading. (Tunghai 04) – L2 corpus
(ii) The only category selected by approximately the same proportion of learners was I practiced listening during in-class activities, suggesting that a similar proportion of learners in both groups felt that the activities they engaged in during their Listening
& Speaking classes were beneficial. (Ohio 36) –Native corpus
(iii) In other words, these learners felt that the presence of the photo freed up attentional resources…(SSLA 86) –Journal corpus
c. Non-factive Reporting Verbs
In native and journal corpus, the verb report is found to be the most frequently used verb that co-occurs with nominalized subjects as in the following examples.
(i) Forty-three percent of the subjects reported that they use AAE frequently (See Table 4). If these students are not aware of the structure and rules, they may not be aware of the usage patterns. (Miami 31) –Native corpus
(ii) Most of these learners (42.1%) reported that when the photo was present, they were able to pay greater attention to the details of the photos and to describe all the activities portrayed…(SSLA 86) –Journal corpus
From the examples above, it is noted that the verb report usually collocates with nouns like subjects and learners, and the verb is generally employed to present the results obtained from the participants or interviews in a study.
82
Comparatively, in L2 corpus, the verbs that tend to co-occur with nominalized human subjects are said, claim, mention etc. Most of these verbs fall under the non-factive reporting verb category. The verbs such as mention and said, which are considered to be colloquial verbs, are found generally associated with the participants or interviewees’ opinions.
(i) Some interviewees mentioned that they preferred cooperative learning more than individual learning or teacher’s instruction for different reasons. (NTUST 1) –L2 corpus
(ii) In addition to the difficulties mentioned above, time was considered a problem. The participants said that they were busy with schoolwork and hence they had limited time for extensive reading outside of classroom. (Tunghai 4) –L2 corpus
Table 4.12 Frequency of Verbs in Nominalized Subjects
L2 Native Journal
Cognitive RF NF Cognitive RF NF Cognitive RF NF
believe 21 2.55 believe 26 3.19 believe 13 0.84
thought 19 2.30 hope 2 0.25 posit 2 0.24
assume 5 0.61 doubt 2 0.25 doubt 2 0.24
hope 4 0.49 suspect 1 0.12
hypothesize 3 0.36 imply 1 0.12
consider 2 0.24 predict 1 0.12
Subtotal 54 6.55 Subtotal 33 4.05 Subtotal 17 1.32
Sensory RF NF Sensory RF NF Sensory RF NF
feel 26 2.42 feel 17 2.08 feel 10 1.20
Subtotal 26 2.42 Subtotal 17 2.08 Subtotal 10 1.20 Non-factive RF NF Non-factive RF NF Non-factive RF NF
said 48 5.82 report 17 2.08 report 17 2.03
claim 39 4.74 state 17 2.08 state 9 1.08
mention 35 4.25 claim 16 2.08 said 7 0.84
state 29 3.52 found 14 1.72 found 7 0.84
found 23 2.79 indicate 13 1.59 claim 7 0.84